environmental sedimentology

Environmental Sedimentology

ESA01 9/5/06 2:58 PM Page i

ESA01 9/5/06 2:58 PM Page ii

Environmental Sedimentology

Edited by Chris Perry and Kevin Taylor

Department of Environmental and Geographical Sciences,Manchester Metropolitan University

ESA01 9/5/06 2:58 PM Page iii

© 2007 by Blackwell Publishing Ltd

BLACKWELL PUBLISHING350 Main Street, Malden, MA 02148-5020, USA9600 Garsington Road, Oxford OX4 2DQ, UK

550 Swanston Street, Carlton, Victoria 3053, Australia

The right of Chris Perry and Kevin Taylor to be identified as the Authors of the Editorial Material in this Work has been asserted in accordance with the UK Copyright,

Designs, and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording

or otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the prior permission of the publisher.

First published 2007 by Blackwell Publishing Ltd

1 2007

Library of Congress Cataloging-in-Publication Data

Environmental sedimentology / edited by Chris Perry and Kevin Taylor.p. cm.

Includes bibliographical references and index.ISBN-13: 978-1-4051-1515-5 (pbk.: acid-free paper)

ISBN-10: 1-4051-1515-7 (pbk.: alk. paper)1. Sedimentology. 2. Geology–Environmental aspects.I. Perry, Chris (Chris T.) II. Taylor, Kevin (Kevin G.)

QE33.E58 2007551.3–dc222006023456

A catalogue record for this title is available from the British Library.

Set in 10/12.5pt Sabonby Graphicraft Limited, Hong Kong

Printed and bound in Singaporeby Fabulous Ltd

The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free

practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

For further information onBlackwell Publishing, visit our website:

www.blackwellpublishing.com

ESA01 9/5/06 2:58 PM Page iv

Contents

List of case studies viList of contributors viiiPreface ix

1 Environmental sedimentology: introduction 1Chris Perry and Kevin Taylor

2 Mountain environments 32Jeff Warburton

3 Fluvial environments 75Karen Hudson-Edwards

4 Lake environments 109Lars Håkanson

5 Arid environments 144Anne Mather

6 Urban environments 190Kevin Taylor

7 Deltaic and estuarine environments 223Peter French

8 Temperate coastal environments 263Andrew Cooper

9 Tropical coastal environments: coral reefs and mangroves 302Chris Perry

10 Continental shelf environments 351Piers Larcombe

References 389Index 428

ESA01 9/5/06 2:58 PM Page v

Case studies

1.1 Anthropogenic modifications to sediment supply, northern Gulf of California 181.2 Temporal changes in sediment composition as a result of pollution: archives of

lead pollution 211.3 Sediment dredging and treatment in the Port of Hamburg, Germany 262.1 Huascarán, Yungay, Peru 1970 522.2 Impact of an extreme tectonic/volcanic event on a mountain sediment system –

Mount St Helens eruption 1980 542.3 Impact of an extreme climate event on the hillslope sediment system – Cyclone Bola,

New Zealand 573.1 Past, present and future impact of the Three Gorges Dam on the Yangtze River, China 953.2 Fluvial responses of rivers to inputs of mine wastes: active transformation and

passive dispersal 1003.3 Styles of deposition of sediment-borne radionuclide contaminants in the

River Techa, Urals, Russia 1024.1 Organic toxins in sediments: Baltic Sea 1334.2 Impact of global change on lake characteristics: Lake Batorino, Belarus 1354.3 Impact of eutrophication/oligotrophication: Lake Miastro, Belarus 1395.1 Influence of tectonics on rates of sediment production, delivery and routing,

the Sorbas Basin, south-east Spain 1715.2 Climate controls on sediment production and delivery, the pluvial lakes of the

Basin and Range Country, American South-west 1745.3 Runoff and sediment movement on a hillslope scale, results of the MEDALUS and

IBERLIM European Union projects 1796.1 Road-deposited sediment, Manchester, UK 1946.2 Sources and dynamics of urban river sediments: Rivers Aire and Calder,

West Yorkshire, UK 1976.3 Sedimentation in an urban water body: Salford Quays, UK 2156.4 Study of road-deposited sediment management: Tampa, Florida, USA 2197.1 Accretion and erosion cyclicity in the Severn Estuary, UK 2397.2 The impact of dams on deltas and estuaries: the Nile delta and the Aswan

High Dam 2497.3 The recycling of contaminants from eroding salt marshes 2538.1 Impact of storms on sediment movement: Texas coast, USA 2808.2 Shoreline changes near Wexford Harbour, Ireland 2858.3 Impacts of reduced fluvial sediment supply to California beaches, USA 2919.1 Cyclical phases of mangrove shoreline accretion and erosion, French Guiana 326

ESA01 9/5/06 2:58 PM Page vi

LIST OF CASE STUDIES vii

9.2 Cyclone-controlled sediment distribution on the Great Barrier Reef shelf (the ‘cyclone pump’ model) 331

9.3 Decadal-scale changes in carbonate sediment compositions across the Florida reef tract, USA 337

9.4 Long-term persistence of oil residues in mangrove sediments, Panama 34110.1 Time-scales of cyclone-driven sedimentary processes on the northern

Great Barrier Reef shelf 35910.2 The south-east Australian and south-east African continental margins:

shelf sedimentation and dispersal influenced by intruding oceanic currents 365

ESA01 9/5/06 2:58 PM Page vii

Contributors

Andrew Cooper School of Biological and Environmental Sciences, University of Ulster,Coleraine Campus, Coleraine, Co. Londonderry BT52 1SA, UK

Peter French Department of Geography, Royal Holloway, University of London,Egham, Surrey TW20 0EX, UK

Lars Håkanson Department of Earth Sciences, Uppsala University, Villav. 16, 752 36 Uppsala, Sweden

Karen Hudson-Edwards School of Earth Sciences, Birkbeck College, University of London, Malet Street, London WC1E 7HX, UK

Piers Larcombe Centre for Environment, Fisheries and Aquaculture (CEFAS), Pakefield Road, Lowestoft, Suffolk NR33 OHT, UK

Anne Mather Department of Geography, University of Plymouth, Plymouth, Devon PL4 8AA, UK

Chris Perry Department of Environmental and Geographical Sciences, Manchester Metropolitan University, John Dalton Building, Chester Street, Manchester M1 5GD, UK

Kevin Taylor Department of Environmental and Geographical Sciences, Manchester Metropolitan University, John Dalton Building, Chester Street, Manchester M1 5GD, UK

Jeff Warburton Department of Geography, University of Durham, Durham DH1 3LE, UK

ESA01 9/5/06 2:58 PM Page viii

Preface

It will be clearly apparent to students of the earthsciences that a number of environmental issueshave gained in importance over the past few years,and that sedimentology has a role to play ininforming the debate about their impacts. Theseissues include recent and future climate change,and a wide range of anthropogenic pressures thatinfluence earth surface systems from a range ofphysical, chemical and ecological perspectives.Many sediment systems act as excellent archivesof past environmental change, allowing us awindow into the recent past, as well as provid-ing tools for monitoring change within activesedimentary environments. These in turn can beused to inform management strategies.

Sedimentology text books traditionally do agood job of introducing the concepts of sedimento-logy and dealing in detail with the principles ofsedimentology. These text books, however, tendto focus primarily on facies analysis and basinprocesses of sediment accumulation, and on theinterpretation of ancient sedimentary environ-ments. Although these cover many aspects ofmodern depositional environments, the emphasisis primarily on using modern systems to inter-pret the geological record. More applied sedi-mentology text books effectively focus on theapplication of sedimentology to the oil and gasindustry. It is increasingly being recognized thata more environmental approach to sedimento-logy is needed. In this context EnvironmentalSedimentology represents a rapidly expandingresearch field, which draws upon both the tradi-tional aspects of sedimentology and the moreapplied areas of the discipline, and also inter-faces with the fields of hydrology, geomorpho-logy, engineering, biogeochemistry and ecology.

Essentially it is concerned with understandingthe development of recent sedimentary systems,and examining their response to both natural andanthropogenically induced disturbance events.

The aims of this text book are:1 to outline the boundaries of the field of Envir-onmental Sedimentology;2 to allow those students with a prior ground-ing of sedimentological principles to developtheir knowledge in the environmental aspects ofthe subject;3 to allow environmental scientists and physicalgeographers access to the field of environmentalsedimentology;4 to allow more specialist groups (e.g. civilengineers, legislators) to gain information onthis topic.

The target audiences for this book are earthscientists, environmental scientists, physicalgeographers, hydrologists and civil engineers.Although we assume that earth scientists willhave a grounding in the principles of sedimento-logy, we recognize that some of the other groupsmay not have. Therefore, the introductory chapter not only summarizes the principles ofenvironmental sedimentology, but also providesan overview of the basics of sedimentology. Weencourage readers of this text to make free refer-ence to the available text books on fundamentalsedimentology where appropriate.

The book is structured such that it outlinesprocesses and issues associated with a range of terrestrial (i.e. upland, fluvial, lake, arid and urban) and coastal and shallow-marine sedimentary environments. We do not incor-porate aspects of groundwater hydrology,which, although often recently described as an

ESA01 9/5/06 2:58 PM Page ix

x PREFACE

environmental sedimentological topic, essen-tially deals with water, rather than sediments.We also do not intend to specifically address thetopic of soils, which are not strictly sediments,but the initial products of bedrock and organicmaterial degradation. Each chapter has a similarstructure and, for each specified environment,examines aspects of: sediment sources and sedi-ment accumulation; the processes and impactsof natural disturbance events; the processes andimpacts of anthropogenic activities; sedimentsystem management and remediation; and issueslikely to be of concern in the future, such as shortto medium term (< 100 years) climatic change,sea-level rise and increased anthropogenic influ-ence. Each chapter is authored by a leader intheir particular field and by the very nature ofthis approach, each chapter should not be con-sidered to be uniform in its format. Each authorhas presented the major environmental pro-cesses and products in a manner in which theyfeel most appropriate, inevitably reflecting thatauthor’s strengths. The reader will, therefore,find some chapters that focus on physical pro-cesses, others on the chemical aspects, and yetothers with a more numerical eye on environ-mental sedimentology. This reflects the widerange of disciplines that inform the subject.

Each chapter provides a wealth of up-to-datereferences to allow the student to follow up on the processes and products discussed in thechapters in more detail. Most, but not all, ofthese references are in primary scientific journalsand reports, but where it is felt advantageous,reference is also made to text books in the field. An important part of each chapter is theinclusion of a number of case studies, which actas self-contained examples to illustrate some of the key concepts discussed in the respectivechapters. Each case study, geographically spe-cific, has additional references pertinent to thatexample.

Numerous individuals have provided inputsinto the development of this book, but in particu-lar we would like to thank the following for theircomments and reviews at various stages: HeleneBurningham, Sue Charlesworth, Ian Drew, SimonHaslett, Piers Larcombe, David Nash, Phil Owens,Laura Shotbolt and Chris Vivian. We wouldalso like to thank Delia Sandford, Ian Francisand Rosie Hayden at Blackwell for help andadvice throughout this project.

Chris Perry and Kevin TaylorManchester Metropolitan University

ESA01 9/5/06 2:58 PM Page x

1.1 ENVIRONMENTAL SEDIMENTOLOGY: DEFINITION AND

SCOPE OF CHAPTERS

Environmental sedimentology represents a rela-tively new subdiscipline of the earth sciencesand, as such, the boundaries of the field are not clearly defined. Herein we define environ-mental sedimentology as . . . ‘the study of theeffects of both man and environmental changeupon active surface sedimentary systems’. Con-sequently, environmental sedimentology can beregarded as the study of how both natural andanthropogenic inputs and events modify theproduction and accumulation of the physicaland biogenic constituents of recent sedimentarydeposits. The field of environmental sedimento-logy has evolved gradually over the past twodecades, largely owing to an increased recogni-tion of the influence that anthropogenic activ-ities are exerting upon sediment production and cycling. Studies in these areas reflect a needto address issues of sedimentological changedriven by environmental or land-use modifica-tion or contamination. This, in turn, has pro-moted increasingly integrated approaches toexamining the dynamics of, and interlinkagesbetween, sedimentary environments, the natureof which can be illustrated by one example,namely studies that link catchment processes(and anthropogenically induced changes in catch-ment sediment yields) with sediment supply to(and through) the coastal zone. Here, disciplinessuch as slope geomorphology, fluvial sedimento-logy, hydrology, coastal and marine sedimento-logy, and coastal management combine to assess

1Environmental sedimentology:

introduction

Chris Perry and Kevin Taylor

interlinked issues of sediment production, trans-port and accumulation.

This book is divided into nine main chapters,each of which deals with a distinct sedimentarysystem. These are delineated as follows; moun-tains, fluvial, arid, lacustrine, urban, temperateintertidal (estuarine and deltaic), temperatecoastal (beach, barrier island and dune), trop-ical coastal (coral reef and mangrove) and con-tinental shelf. Although this structure provides a convenient approach for distinguishing anddescribing individual sedimentary environments,such subdivisions are to some extent arbitraryand, in light of the sediment exchanges that occurbetween environments, not necessarily ideal. Asa result, the linkages that exist between envir-onments are highlighted where appropriate. Inaddition, the book does not set out to reviewevery type of sedimentary environment, but ratherto provide a framework for the subject in thecontext of a number of key sedimentary systemsand settings. Given these caveats, each of thechapters examines aspects of sediment supplyand accumulation, the response of the individualsedimentary systems to natural and anthropogenicchange, and issues of sediment management andremediation, and reviews the potential responsesof these sedimentary systems to issues such asclimatic and environmental change.

The chapters in this book essentially deal with sedimentological processes and geomor-phological changes that have been operatingover short-to-medium (< 100 yr) time-scales(biological time-scales of Spencer 1995; seculartime-scales of Udvardy 1981), although many of

ESC01 9/8/06 2:39 PM Page 1

2 CHRIS PERRY AND KEVIN TAYLOR

the sedimentary landforms discussed also inevit-ably represent the products of longer-term sedi-ment accumulation. Hence, in each chapter thesemay encompass the daily processes of sedimenttransport and reworking, through to the effectsof rapid climatic and sea-level change (Fig. 1.1).In this context, short-term changes (< 1 yr) withina coastal sedimentary environment may, forexample, include the processes of tidal erosionand deposition, seasonal changes in beach pro-files or morphological change due to storm events;in the short-to-medium term (< 10 yr), massmovement of unstable cliffs, spit progradation,or barrier breaching; and in the medium-term(up to 100 yr), coastal landform progradation,changes in delta morphology, or coastal retreat.Superimposed upon these processes may be arange of anthropogenic activities (e.g. sea-wallconstruction, sand dredging or contaminantinputs), which may influence not only the dy-namics of the sedimentary system, but also theassociated floral and faunal components of thesystem. Many of these biological components(e.g. dune plants, mangroves or corals) are oftenof sedimentological significance in their ownright, either as sources of sediment or as agentsof sediment trapping and stabilization.

This introductory chapter reviews the funda-mentals of sediment production and the principlesof sediment transport and deposition. In addi-tion, consideration is given to the types of issuesthat are discussed in the respective chapters, andwhich help to define the discipline of environ-

mental sedimentology. This chapter also outlinesthe magnitude and frequency of predicted shiftsin global climatic and environmental conditionsthat have relevance to the development of bothterrestrial and marine sedimentary systems overthe coming century. These include issues such assea-level change, global warming and changes intemperature, precipitation and storm frequency,and thus provide a framework for discussionwithin the respective chapters.

1.2 SEDIMENT PRODUCTION AND SUPPLY WITHIN

SEDIMENTARY ENVIRONMENTS

The composition of sediments that accumulatewithin individual sedimentary environments isprimarily a reflection of three main factors:1 the sediment source;2 the processes of sediment transport and deposition, which determine whether sedimentis retained or transported through a specificenvironment (these mechanisms are outlined insection 1.3);3 the chemical processes operating within thesediment or water column, for example, car-bonate and evaporite precipitation, chemicaldiagenesis.In terms of the initial supply of sediment into asedimentary system, three basic sediment typescan be delineated. These are: (i) detrital minerals,(ii) biogenic or organic sediments and (iii) anthro-pogenic particles and compounds.

Time-scale (years)

Climatic andsea-level change

Spit/barrier retreator progradation

Mass movementof cliffs

Storm reworkingand erosion

Changes inbeach profile

Tidal erosionand deposition

100,00010,0001,0001001010.10.01

Fig. 1.1 Time-scalesover which differentgeomorphologicalprocesses operate withina hypothetical coastalenvironment. (Adaptedfrom Woodroffe 1992.)

ESC01 9/8/06 2:39 PM Page 2

ENVIRONMENTAL SEDIMENTOLOGY: INTRODUCTION 3

1.2.1 Detrital minerals

Detrital minerals, such as quartz and feldspar,along with heavy minerals, form a primary component of many terrestrial and marine sediments. These minerals are initially releasedby weathering processes and are progressivelyeroded and transported into, and through, arange of sedimentary environments. As a result,initial mineralogical composition of the bedrockoften influences the relative abundance of theindividual minerals that are released. This con-trol is clearly illustrated in studies of suspendedsediment compositions within river catchmentswhere individual tributaries are underlain bybedrock of differing geological compositions. Inthe River Ouse catchment (north-east England),for example, individual tributaries drain areasof differing geology (Carboniferous, Permian/Triassic and Jurassic). Suspended sediments in

the different tributaries have distinct mineralog-ical and magnetic signatures that demonstratevariations in the relative importance of differ-ent rock units as sources for fluvial sediment(Fig. 1.2). Local variations are attributed tovariations in the rates of erosion and sedimentsupply from the different geological units(Walling et al. 1999).

In reality, these detrital minerals rarelyundergo a simple source to sink transport route,but instead are subject to numerous phases of weathering, transport, deposition, storage,lithification, reworking and redeposition. Forexample, detrital sands in the Orinoco drainage-basin of South America are derived from similarbedrock material, but the nature of relief andchemical weathering markedly alter the graincomposition (Johnsson et al. 1991). Materialderived from steep, orogenically-active terrainsundergoes limited chemical weathering, whereas

Hawes

York

River Swale

River Ure

River Nidd

River Wharfe

River Ouse

20 km

N

Carboniferous

Permian & Triassic

Jurassic

Fig. 1.2 Composition of suspended sedimentsamples recovered fromthe River Ouse (UK) andfour of its tributaries thathave different underlyinggeologies. (Adapted fromWalling et al. 1999.)

ESC01 9/8/06 2:39 PM Page 3


in low-relief parts of the catchment thick soilsaccumulate and the detrital grains becomehighly altered chemically. Furthermore, sedi-ment deposited in low-lying alluvial floodplainsundergoes additional chemical weathering andalteration, further modifying their composition.Consequently, within any individual environ-ment sediments tend to be derived from a rangeof source areas and weathering regimes, and theseboth contribute to, and influence, the composi-tion of the accumulating sediment (Fig. 1.3).Over the time-scales considered in this book, therelease of sediment from previously depositedsedimentary sequences should, therefore, alsobe regarded as a key sediment source.

1.2.2 Biogenic and organic sediments

In addition to detrital minerals, significantamounts of sediment are derived from theremains of skeletal carbonate-secreting organ-isms. These form across a wide range of marineenvironments (Schlager 2003), although markedlatitudinal variations occur both in the typesand rates of biogenic sediment production (Lees1975; Carannante et al. 1988). Such productionpeaks in the low latitudes and, in particular, in the vicinity of coral reefs, where carbonatesediments often represent the primary sedi-ment constituents (see Chapter 9). Even in theseenvironments, however, marked variations inthe composition of sediment assemblages are

evident across individual reef or platform envir-onments and reflect subtle spatial variations inmarine environmental conditions (e.g. light andwave energy). These influence the compositionof the reef community and, hence, the com-position of the sediment substrate. In high-latitude settings, biogenic carbonate productionmay remain important but in many settings isvolumetrically ‘swamped’ by terrestrial inputsof detrital minerals.

Carbonate deposits are also relatively commonwithin a range of intertidal, terrestrial and fresh-water settings, and are associated primarily withphysico-chemically induced carbonate deposition.On a localized scale these represent importantsources of carbonate to the sediment record. Forexample, within intertidal and supratidal settingsthat are characterized by high aridity and highevaporation rates, the precipitation of evaporiteminerals such as gypsum (CaSO4.2H2O) andanhydrite (CaSO4) is common. At present,extensive evaporite-rich sediments occur in thesouth-east Arabian Gulf and an excellent reviewof their occurrence is given by Alsharhan &Kendall (2003). The hydrology of such deposi-tional settings has been reviewed by Yechieli &Wood (2002).

In some freshwater fluvial and lacustrine set-tings carbonate deposition also occurs and canlead to the development of significant carbonatebodies (in some cases these have been describedas forming ‘freshwater reefs’; Pedley 1992).

River/estuarinesupply

LongshoretransportOnshore/

offshoretransport

Erosion of pre-existing sediment deposits

Zone of deposition

Mixing/reworking - physical- biological

In situ (biogenic)sediment production

Anthropogenic inputs/outputs- contaminants- sediment extraction- sediment dumping

Fig. 1.3 Schematicdiagram illustrating thedifferent potential naturaland anthropogenicsources of sediment into a nearshore marineenvironment.

ESC01 9/8/06 2:39 PM Page 4


Deposition is driven by a range of physico-chemical and biologically mediated processes,the latter in association with microbial mats thatfacilitate local reductions in CO2 and thus car-bonate precipitation (Pedley 2000). These depositsare termed tufa where deposition occurs underambient environmental conditions, or travertinewhere the deposits are associated with thermalactivity (Ford & Pedley 1996) and have beenshown to have potential as recorders of palaeo-climatic information (Andrews et al. 1994).

Organic inputs, derived from plant material,can also contribute abundant material to sedi-ment substrates. This is particularly the case insalt marshes (see Chapter 7) and mangroves (seeChapter 9). Along mangrove-colonized shorelines,where external inputs of sediment (siliciclasticor carbonate) are minimal, this organic mater-ial can be the main substrate contributor andleads to the development of mangrove peat(Woodroffe 1983). Biogenic, but non-carbonatesediment, contributors such as diatoms are alsoimportant within, for example, lake environ-ments (Chapter 4). The progressive accumula-tion of such microfossils within lake sedimentshas proved to be an effective long-term recorderof a range of environmental parameters such as effective moisture, i.e. lake water levels, and of temperature (Battarbee 2000) and have thusbeen widely used as proxy records of climaticand hydrological change (e.g. Bradbury 1997).

1.2.3 Anthropogenic particles and compounds

Increasingly important in many sedimentarysystems are inputs of anthropogenically sourcedsediments. These include both sediment grainsthat come from material that is anthropogenicin origin (e.g. building material, industry) andsedimentary materials that have been heavilyimpacted by anthropogenic activity. A goodexample of anthropogenic-derived sediment isthat present within urban environments (seeChapter 6). In this environment, as well as soiland vegetation sources, sediment is sourcedfrom vehicle wear, building material, combus-tion particles and industrial material. All of this material has chemical and mineralogical

properties distinct from natural sediment grainsand, as a consequence, interacts with the envir-onment in a different manner.

Another significant component of modernsediment, mostly absent from pre-industrial agesediments, are contaminants. A contaminantis commonly defined as a substance releasedinto the environment without a known impact(Farmer 1997), or the presence of elevated con-centrations of substances in water, sediments ororganisms (GESAMP 1982). In neither of thesedefinitions is the potential to cause environmentalharm attributed to a contaminant. This is in con-trast to a pollutant, which is more specificallydefined as a substance that either causes harm to the environment or exceeds an environmentalstandard. Contaminants in sediments take a vari-ety of forms, including metals, inorganic elements,nutrients, organic compounds and radionuclides,and the major sources of these contaminants arehighlighted in Chapter 3 (Table 3.2). It is import-ant to be aware that many of these contaminantscan be sourced from natural processes as well asanthropogenic activities, although in most casesanthropogenic inputs tend to dominate.

Contaminant sources to sediments may be of particulate, dissolved or gaseous form, butfor most contaminants the particulate form is dominant (Horowitz 1991). Although con-taminant sources are predominantly particulate,there are important exceptions. Contaminantsfrom sewage treatment works (e.g. Zn and P)can be predominantly in solute form, and metalsfrom acid mine drainage are also in dissolvedform at source owing to the low pH of thewaters. These dissolved contaminants, however,commonly become associated with the sedi-ment phase, via mineral precipitation or surfaceadsorption, as solution concentration, pH andEh change through mixing with dilute river water(Boult et al. 1994).

Contaminant sources may take one of twogeneral forms, point source and diffuse (non-point) source. Point sources of pollution origin-ate from a single location and include mines andmine waste, landfill sites, factories, waste watertreatment works and bedrock mineralization.Diffuse sources of contaminants originate from

ESC01 9/8/06 2:39 PM Page 5


a wide area and can be defined as ‘pollution arising from land-use activities (urban andrural) that are dispersed across a catchment orsubcatchment, and do not arise as a process ofindustrial effluent, municipal sewage effluent,deep mine or farm effluent discharge’ (Novotny2003). Examples include direct atmosphericdeposition, urban runoff and sediments (i.e.from the road network), agricultural runoff andsediments (i.e. from soil erosion), the reworkingof floodplain sediments (i.e. by bank erosion)and background geology.

1.2.4 Particle description and classification

Regardless of origin, individual sedimentaryparticles are typically described in terms of theirgrain size and shape. Grain size is an importantparameter both from a descriptive perspectiveand in relation to understanding sediment trans-port and deposition (see section 1.3). For largerparticles, measurements of three orthogonal axes

are typically made and are used to calculate amean diameter. For smaller particles, grain sizeis typically determined by grading the samplesthrough a set of sieves (see McManus 1988). Anumber of schemes have been devised to describeand measure grain size, but one of the mostwidely used is the Udden–Wentworth scheme(Fig. 1.4a).

Descriptions of sediment shape are somewhatmore complex and may be taken to compriseelements of a particle’s form, roundness and texture. Roundness is usually described on thebasis of comparisons with visual identificationcharts. Form is also usually quantified by describ-ing grains in terms of one of four standard classes:oblate, equant, bladed or prolate, which reflectthe relationship between the short, intermediateand long axes of grains (Fig. 1.4b). Other usefulschemes combine elements of both roundnessand sphericity in visual comparison charts (e.g.Powers 1982). Particle sorting describes the rangeof grain sizes that occur within a sedimentary

Phi (φ) mm μm

−7 −6 −5 −4 −3 −2 −1 0 1234 5 6 7 8 9

Cobbles

Pebbles

GravelV. coarseCoarseMediumFineV. fine

Silt

Clay

San

d

Udden–Wentworth

12864321684210.5 500

250125633115.67.83.92.0

(a) (b)

Oblate (disc, tabular) Equant(spherical)

Bladed Prolate (roller,

rod-like)

0.8

0.6

0.4

0.2

dIdL

dSdI

0.2 0.4 0.6 0.8

dS - shortest axisdI - intermediate axisdL - longest axis

Very wellsorted

Wellsorted

Moderatelysorted

Poorlysorted

Very poorlysorted1 3 5 7

(c)

Fig. 1.4 (a) The Udden–Wentworth scheme, which is widely used to describe grain-size categories. (b) Classification scheme used to describe particle form based on the ratio between the short, intermediate and longest grain axes. (Adapted from Graham 1982.) (c) Visual comparison chart used for describing the degree of sediment sorting within a sediment deposit. (Adapted from Graham 1982.)

ESC01 9/8/06 2:39 PM Page 6


deposit (Fig. 1.4c) and can be calculated by measuring the dispersion of grain size aroundthe mean. This is again a useful parameter as itcan be used, along with grain size data, to inferinformation about the environments of sedimentdeposition and the history of sediment rework-ing (e.g. McManus 1988).

1.3 MECHANISMS OF SEDIMENT TRANSPORT AND

ACCUMULATION

The transport and deposition of sediment withinand through different sedimentary environmentsmay occur within a variety of mediums (water,wind or ice), and the thresholds for sedimententrainment and transport represent a funda-mental control on both the character and develop-ment of specific sedimentary deposits, as well astheir response to fluctuating energy regimes.The classic work of Hjulström (1935) demon-strated the relationship that exists between thevelocity of fluid flow and the size (diameter) ofsediment that can be moved within a fluid. At itsmost simplistic this demonstrates that sedimentwill be deposited when flow rates drop belowthe fall velocity for a particle of a given size.However, the relationship between these twoparameters is non-linear so that, for example,much higher flow velocities are required toentrain highly cohesive fine clay and silt-richsediments (Fig. 1.5). Although in reality theseentrainment/transport thresholds vary from this model depending upon sediment substrateand individual grain characteristics (e.g. shape,structure, density) as well as the flow charac-teristics of the fluid medium, a basic grain-size–flow-velocity relationship is demonstrated thatcan be broadly applied within both fluvial andmarine settings. This section outlines some of thekey physical parameters that control sedimententrainment, transport and settling (deposition),and highlights the main sedimentary processesthat operate within the different sedimentaryenvironments discussed in this book. For furtherdetails about the physics of sediment transportand deposition reference should be made to textssuch as Allen (1985) or Leeder (1999).

1.3.1 Sediment entrainment

The entrainment of sediment by a fluid (mostcommonly water or wind), and thus the poten-tial for sediment transport, is determined by therelationship between (i) fluid density (which isthe weight per unit volume of a fluid – usuallyexpressed as ‘specific gravity’), (ii) fluid viscosity(the resistance of a fluid to deformation or flow– this is measured as a ratio between the shearstress and the rate of deformation) and (iii) thevelocity of fluid flow. These parameters exert aninfluence on the nature of the flow regime withina fluid medium and, in particular, determinewhether flow that occurs immediately above thesediment substrate (within the boundary layer)is laminar or turbulent. In situations charac-terized by laminar flow, flow streamlines runparallel to the substrate, flow velocity is low and viscosity is high (Fig. 1.6a). Under turbulentflow, the streamlines move in a series of randomeddies, flow velocity is high and viscosity is low(Fig. 1.6a). The threshold between these twoflow states is expressed by the Reynolds number(R), which describes the ratio between meanvelocity over a defined distance or depth, and

100.0

10.0

1.0

0.1Log

velo

city

(cm

s−1

)

0.01 0.1 1.0 10 100 Log grain size (mm)

DepositionTransportation

Erosion

Fallvelocity

Clay Silt Sand Pebble

Grain size (φ)

8 6 4 2 0 −2 −4 −6

Gravel

Entrainment velocity

Fig. 1.5 Hjulström’s (1935) graph showing the relationshipbetween flow velocities and sediment grain size and thecorresponding fields in which erosion, transport and deposition occur.

ESC01 9/8/06 2:39 PM Page 7


fluid viscosity, and is determined by the follow-ing equation:

R = Udp/μ

where U is the particle velocity, d is the particlediameter, p is the particle density and μ is the fluidviscosity. In the context of particles moving in afluid, a low Reynolds number (< 500) describesfluid flow occurring in a laminar fashion, whereasat high Reynolds numbers (> 2000), fluid flow is turbulent. Between these two values flow isdescribed as transitional (Allen 1985). Flow turbulence increases both proportionally withvelocity and as bed surface roughness increases.

Another important coefficient in terms of fluiddynamics (the Froude number) explains the ratiobetween the force required to stop a movingparticle and the force of gravity. Within openchannels the Froude number (F) is determined as:

F = U/√gD

where U is the average current velocity, g isacceleration due to gravity and D is the depth ofthe channel. As flow velocity increases the Froudenumber approaches 1, a value that separateslower (< 1) flow regimes from higher (> 1) flowregimes (Allen 1985). Flow velocities increasefrom lower to upper flow and associated withthis is an increase in the amount and size of sediment that can be entrained and transported.This, in turn, will influence the structure of thesedimentary bedforms that develop (Fig. 1.6band section 1.4.1).

A number of forces act upon a sediment lyingon a substrate surface (Fig. 1.7a), and theseinfluence the potential for entrainment. The keyforce here is bed shear stress, which is related to the velocity of flow. This represents the forceacting per unit area parallel to the bed andexerts a fluid drag across the grain. If this dragexceeds the frictional and gravitational forcesacting on the grain, then lift and entrainmentwill occur. Sediment entrainment thresholds thus

Flo

w r

egim

e

Low

er

U

pper

Increasing flow velocity

Froude number > 1

Froude number < 1

Ripples

Dunes

Plane bed

Antidunes

(a)

(b)

Laminar Transitional Turbulent

U (y)

U (y)

U (y) = flow velocity (m s−1) parallel to substrate

Edge of boundary layer

Fig. 1.6 (a) The nature of flow regimeswithin a fluid medium. During laminarflow, the flow streamlines run parallel tothe substrate, flow velocity is low andviscosity high. Under turbulent flow, thestreamlines move in a series of randomeddies, flow velocity is high and viscositylow. Intermediate flow is described asbeing transitional. (Adapted from Allen1985.) (b) The relationship between flow regime and sediment bedformdevelopment. As flow velocity increasesfrom lower to upper flow, the amount andsize of sediment that can be entrained and transported increase, leading to achange in sediment bedform structure.(Adapted from Selley 1994.)

ESC01 9/8/06 2:39 PM Page 8


occur at a critical shear velocity, the value ofwhich varies with sediment grain size (althoughthis is complicated by substrate specific vari-ations in grain size, particle density, sedimentpacking and grain imbrication). The interrela-tionship between these variables is highlightedin the Hjulström graph (Fig. 1.5), which illus-trates a general (and fairly intuitive) rule wherebyincreasing flow velocities are required to entrainincreasingly larger sized particles. This rule, how-ever, breaks down where the substrate is domin-ated by very fine sands, silts and clays because of the cohesive nature of such material. In suchcases, much higher flow velocities are requiredto entrain particles and this helps to explain whyfine silts and clays can accumulate within tidallyinfluenced estuarine and deltaic environments(see Chapter 7). Velocity–entrainment relation-ships are also complicated as fluid moves acrossa sediment substrate because grains protrudingfrom the substrate cause flow to be constricted.This causes the streamlines above the grain toaccelerate and thus to exert a fluid lift (Fig. 1.7a).Within aeolian environments, wind velocitiesthat exceed the critical shear stress for specificsediment grain sizes are also required for entrain-

ment, although they occur at higher velocitiesthan in water.

1.3.2 Sediment transport

Once entrained, sediment movement occurs inthree ways:1 as bedload material that is too heavy to belifted up into the water and moves by rollingalong the substrate (Fig. 1.7b);2 via the process of saltation whereby lightergrains are temporarily lifted into the fluid andthen settle out (in water, saltating grains typic-ally exhibit short, flat trajectories due to thecushioning effects of fluid viscosity, whereas in air the trajectories tend to be steeper andlonger; Fig. 1.7b);3 in suspension where the lightest particles areheld within the fluid and moved, often in anerratic path (Fig. 1.7b).For sediments of a given grain size these transportmechanisms occur along a gradient of increas-ing flow velocity. Consequently, for a sedimentdeposit comprising a mix of grain sizes it followsthat the occurrence of different sediment sizefractions can be attributed to different transport

(a)Flow direction

ab

c

FL

FD

FW FC

FL Fluid liftFD Fluid dragFC Cohesion/frictionFW Immersed weight

(b)

(c)

a b c = increasing flow velocity where grains are of uniform size or decreasing grain size under conditions of uniform flow

Fig. 1.7 (a) Schematic diagram illustrating the main forces acting on a sediment particle within a moving fluid medium. (Adapted from Allen 1985.) (b) The processes of sediment movement within flowing water; a, rolling; b, saltation; c, suspension. (c) Grain transport due to saltation under conditions of aeolian transport. Grains typically exhibit steeper and longer trajectories during aeolian transport.

ESC01 9/8/06 2:39 PM Page 9


mechanisms. In aqueous environments, solutetransport can sometimes also be an importantmedium for the transport of contaminants (seeChapters 3 and 6). These transport mechanisms,which operate in a range of fluid mediums andin different environments, produce very differenttypes of sedimentary deposits. These are discussedbelow in the context of (i) aqueous environ-ments, (ii) aeolian environments and (iii) glacialenvironments. Consideration is also given to themovement of material by gravitational processes.

1.3.3 Sediment settling

As with sediment entrainment and transport,the major controls on sediment deposition relateto grain size and flow velocity. The rate at which sediment settles (the settling velocity W)within a fluid of a given density is determined by Stokes law:

W = [(P1 − P)g/18μ] d2

where (P1 − P) is the density difference betweenthe fluid and particle, g is the acceleration due to gravity, μ is the fluid viscosity and d is thegrain diameter. The law states that the settlingvelocity of a spherical particle is related both toits diameter and to the difference between thedensity of the particle and that of the surround-ing fluid. In simple terms this means that largersediment grains will settle faster than smallergrains providing they are of equal density. Withinmost sedimentary systems this process is com-plicated, however, by three factors:1 the fact that few grains are completely spherical;2 the fact that grains are often continually incontact within one another and hence disruptsettling;3 the fact that different minerals have differ-ent densities (e.g. quartz 2.65 g cm−3, feldspars2.55–2.76 g cm−3, biotite 2.80–3.40 g cm−3; Allen1985).Hence, in the case of terrigenous sands, whichare commonly dominated by quartz, but withvariable amounts of other detrital and heavyminerals, the different particles have different

settling velocities and thus different transportand settling thresholds. Differences in settlingvelocities are even more complex in systemsdominated by skeletal carbonates because thegrains not only have very different internalskeletal structures (and hence densities), but alsovery different morphologies (see Chapter 9).

1.4 SEDIMENT TRANSPORT IN DIFFERENT SEDIMENTARY

ENVIRONMENTS

1.4.1 Sediment transport in aqueousenvironments

Sediment transport within aqueous environ-ments occurs primarily in association with eithertraction or turbidity currents. Within tractioncurrents the primary mechanisms of sedimentmovement are rolling and saltation (Fig. 1.7b),whereas within density currents transport is asso-ciated both with traction and suspension. Con-sideration is given first to traction currents, whichare important within both fluvial and shallowmarine environments. Within fluvial systems,current flow is nearly always unidirectional andprogressive reworking of fine sediment commonlyleads to the finest material being transportedfurthest downstream. Hence, fluvial systems aretypically characterized by downstream reductionsin mean grain size (see Chapter 3). In contrast,within nearshore settings and in the marine-influenced lower reaches of rivers, currents tendto be bi-directional (owing to the variable flood-and ebb-tide influence) and hence sediments willbe reworked both on- and offshore during thetidal cycle (see Chapters 7 and 8).

Studies of flow regimes under unidirectionalconditions have highlighted clear changes insediment transport mechanisms and sediment-ary structures associated with different currentvelocities. As flow increases, the critical velocitiesrequired to entrain sediment particles are reachedand at this stage sediment starts to move byrolling and saltation. This leads to the develop-ment of ripples and, at slightly higher velocities,dunes (Fig. 1.6b). Such structures are associatedwith lower flow regimes (Froude numbers < 1).

ESC01 9/8/06 2:40 PM Page 10


As flow velocities increase, upper flow regimes(Froude number > 1), characterized by turbulentflow, are reached and under these conditions thesediment bedforms are initially smoothed out to form planar beds and eventually antiduneswhich may migrate upstream (Fig. 1.6b). As flowreduces, a reverse sequence is followed. Hencethrough cycles of river flooding, the mechanismsand processes of sediment movement changewith flow velocity (see Chapter 3).

In contrast, shallow marine environments arecharacterized by bi-directional flow, althoughthe magnitude and frequency of the flood- versusebb-tidal phase varies depending upon localtidal regime and nearshore geomorphology. The potential for sediment transport changesthrough each tidal cycle as flow velocity in-creases through either the ebb- or flood-tidephase and then decreases approaching either lowor high tide (‘slack’ water). In settings where thetide cycle is symmetrical (Fig. 1.8a) sedimentwill be reworked first seaward and then land-ward, but there will be no net transport in eitherdirection. It is more common, however, for tidalcycles to be asymmetrical (Fig. 1.8b), and underthese conditions there will be a net sedimenttransport direction. This situation is common inmany estuaries where fluvial outflow exerts aninfluence on the tidal cycle (see Chapter 7), or inmangrove settings where strong ebb-tide flowsin the mangrove creeks can occur owing to the

frictional effects of high vegetation cover on themangrove flats (see Chapter 9).

Sediment transport is also initiated in near-shore (marine or lacustrine) environments wherewave-generated water motion interacts with the shoreline substrate. Waves are generated bythe frictional effects of wind and this initiateswater particle motion within the upper part ofthe water column. The orbital particle motiondecreases with depth (Fig. 1.9a) until it reacheseffective wave base (defined as half the wave-length), below which there is no wave-inducedwater motion. In open, deeper water, watermotion therefore exerts no influence on sea-floorsubstrates, but as the water shallows nearshorethe oscillating water particles start to interactwith the sea-bed (Fig. 1.9b). As this occurs, thewater particles move in an increasingly ellip-soidal fashion and initiate on- and offshoremovement of sediment.

Transport in aqueous fluids may also occur indensity currents. These are associated both withtraction and suspension transport, and occur dueto density differences between two fluid bodies.Within aqueous environments, density differencescommonly result from variations in temperature,salinity or suspended sediment load where twobodies of water meet. Where the fluid entering abody of water has a higher density, for examplewhere sediment-laden water enters a lake, thedenser fluid will flow beneath the less dense fluid

Time

Lowtide

Hightide

Critical velocityfor sediment transport

Sediment transport ‘window’

(a) (b)

0 0

Ebb

vel

ocity

Flo

w v

eloc

ity

Ebb

vel

ocity

Flo

w v

eloc

ity

Lowtide

Hightide

Time

Critical velocityfor sediment transport

Fig. 1.8 Graphs showing changes in ebb- and flood-tide velocities within (a) symmetrical and (b) asymmetrical tide-cycle settings.Within each phase of the tidal cycle sediment transport occurs only when velocities exceed the critical thresholds for transport. Insymmetrical settings, sediment is moved back and forth but there is no net sediment transport direction. In asymmetrical settings, astronger ebb- or flood-tide phase may result in a net direction of sediment transport.

ESC01 9/8/06 2:40 PM Page 11


and create a density current (Fig. 1.10a; seeChapter 4). Where the fluid entering the body ofwater has a lower density, for example wherefreshwater enters the sea, flow will typically occuras a plume across the water surface (Fig. 1.10b).In the former case, a specific type of density cur-rent, known as a turbidity current, is commonlygenerated. These are capable of transportingvery large volumes of sediment across even very

low slope angles and are thought to be a majorcause of sediment distribution across contin-ental shelf (see Chapter 10) and slope settings.In low-density flows, most of the fine sediment is transported in suspension. This mechanism of transport is common at the distal ends of turbidite deposits where the finest sediment hasremained in suspension, and along the distalmargins of deltas where fine suspended sedimentsettles out along the delta front (see Chapter 7).Settlement of fine grained suspended sediment isenhanced where mixing of fresh and saltwateroccurs. Under these conditions, even slight in-creases in salinity (> 1) will promote the aggrega-tion of fine clay particles. This process is knownas flocculation and leads to an increase in grainsize and thus in grain settling velocity. Floccula-tion is a common process in estuarine environ-ments (see Chapters 7 and 9) and in salt marshes,and may lead to the development of zones ofhigh turbidity and fine sediment deposition.

The largest proportion of the contaminantload in sediment systems is transported by theparticulate matter. For example, Gibbs (1977)suggested that up to 90% of the metal load istransported by sediments in rivers, but this canvary from metal to metal. Similar observationshave been made for organic contaminants, suchas chlorinated organic compounds. This particu-late portion of the contaminant load comprisescontaminant-rich grains (e.g. metal sulphidegrains from tailings effluent) or contaminant

(1) Wind motion over the sea-surface generates sea waves

(2) Sea waves inducean oscillatory motion within water particles

(3) Orbital water motion decreases with depth

Effective wave base (D)(D = L/2)

Wave length (L)

(a) Open water setting (b) Nearshore setting

(4) As waves approach the shore water shallows below wave base and water motion leads to interaction with sea-floor sediments

(5) Shallowing leads to anincrease in wave height. Wave break when wave height equals water depth

Trough

Crest

Waveheight (H)

Fig. 1.9 Schematic diagram illustrating wave motion in (a) open-water settings, and (b) as waves approach the shoreline.

D1 D2

D1 D2

(a)

(b)

High density currents D1 > D2

Low density currents D1 < D2

Turbidite current

Slow settling of sedimentin deeper water

Fig. 1.10 Schematic diagrams illustrating differences in densityflows as fluid enters a standing body of water. (a) High densityflows occur where the entering fluid has a higher density thanthe standing water body. (b) Low density flows occur when thesituation is reversed. (Adapted from Selley 1994.)

ESC01 9/8/06 2:40 PM Page 12


element-bearing Fe and Mn oxide coatings onother particles. Some metals, however, especiallyunder low pH conditions, can be transported insolution (see Chapter 3). This dissolved portionencompasses contaminants that are either trulydissolved, or in colloid form. The partitioning of contaminants between the dissolved and par-ticulate load in aquatic systems depends on bothphysical and chemical factors, including pH,redox, sediment mineralogy, sediment texture,suspended sediment concentration and sedimentgrain size. Grain size is possibly the most signific-ant factor controlling the concentration andretention of contaminants in both suspendedand bottom sediment. Metals in particular havebeen shown to be enriched in the fine silt andclay fractions of sediments, as a result of theirlarge surface area, organic and clay contents,surface charge and cation exchange capacity(see Chapters 3 and 6).

1.4.2 Sediment transport in aeolian environments

Sediment transport within aeolian settingsoccurs primarily associated with either tractioncarpets or in suspension and is common in threemain environments:1 arid deserts;2 associated with shoreward areas of beachesand barrier islands;3 developed around ice caps.Sediment transport in deserts (Chapter 5) andcoastal dunes (Chapter 8) is associated primarilywith rolling and saltation of grains in the tractioncarpet. As in aqueous environments, the criticalvelocities required to entrain sediment increasewith grain size, and high velocities are requiredto entrain fine silt and clay-sized material. Onceentrained, however, such sediment may be trans-ported long distances as dust clouds, a mechan-ism that is known, for example, to transportlarge volumes of Saharan dust across the Atlanticand into the eastern Caribbean (Prospero et al.1970). Controls on the development of aeoliansediment bedforms are discussed in Chapter 5,but are strongly influenced by wind directionand its variability, and by the rate of sedimentsupply. Aeolian sediment deposition around ice

caps is thought to be primarily associated withsuspension transport and these silica-rich sandsare commonly termed ‘loess’.

1.4.3 Sediment transport in glacial environments

Sediment transport within glacial environmentsoccurs associated with a range of transport pro-cesses, which include suspension, aqueous suspen-sion and aqueous traction currents. These formdifferent types of deposits and are associated withdifferent glacial environments. Suspension trans-port, as outlined above, results in the depositionof loess deposits in glacial marginal areas. Aque-ous suspension is associated with the depositionof fine, laminated clay sequences (varves), whereasaqueous traction currents are responsible forextensive fluvioglacial sand and gravel trans-port, and the development of extensive outwashplains (see Chapter 2). Glacial ice also acts as an important sediment transport medium and,although the movement of ice is slow, it isresponsible for significant erosion of underlyingbedrock and sediment. The resultant debris istransported under and within the ice, and isdeposited either along the flanks of glaciers or at the terminal end after the ice starts to melt.These deposits are typically structureless andcomprised of poorly sorted boulder to clay-sizedmaterial. The descriptive term for the sedimentis diamict and when deposited directly by glacierice is termed till. Consequently till is a majorcomponent of glacial landforms such as morainesand drumlins. High-magnitude sediment trans-port events in glacial environments can occurassociated with jökulhlaups – a flood caused by the sudden drainage of a subglacial or ice-dammed lake, commonly triggered by a volcaniceruption (see Chapter 2). These events can trans-port huge volumes of sediment, and result inextensive deposition of outwash deposits.

1.4.4 Sediment transport associated withgravitational processes

Within each of the settings described, an addi-tional agent of sediment transport is gravity andthree main categories of gravitational sediment

ESC01 9/8/06 2:40 PM Page 13


transport are recognized: (i) rockfalls, (ii) slidesand slumps, and (iii) mass flows. These are re-cognized to occur along a continuum wherebythere is an increase in the degree of internal dis-aggregation and a reduction in the concentrationof the sedimentary material (Fig. 1.11). Rock fallsare defined as the collapse of rock or sedimentprimarily along a vertical plane. They may becaused by tectonic movement or by weatheringin upland settings and typically produce screedeposits. Slides and slumps occur over lowerangled slopes and involve transport along aninclined shear plane. They are thus characterizedby movement over both vertical and horizontaldisplacement planes. In slides, the sediment gen-erally remains undisturbed, whereas in slumps the original sedimentary structures are normallydisrupted or destroyed. The presence of wateralong a shear plane acts as a medium to initiateboth slumping and sliding. At higher water contents the process of slumping grades intothat of mass flows, a term used to encompass aspectrum of transport processes including debrisflows and grain flows. Debris flows involve thetransport of rock and fine sediment that ‘flows’downslope as a chaotic mass and these occur ina range of environments from deserts to contin-ental slopes. They typically require the presenceof unconsolidated sediment and steep slopesand, on land, low vegetation cover and heavyrainfall to initiate movement (Chapter 5). Grainflows occur within finer sediments and require

steep slopes and a confined channel margin. Theyoccur most commonly on the continental slopesand form graded deposits.

1.5 POST-DEPOSITIONAL PROCESSES

Processes acting internally and externally upona sediment after deposition can be physical,chemical or biological. Physical processes includecompaction, resuspension, erosion or dredgingof sediment. Chemical and biological processesinclude the series of early diagenetic, bacteriallymediated redox reactions, which result in theoxidation of carbon species (organic matter) andthe reduction of an oxidized species. Althoughpost-depositional processes acting upon sedi-ments are varied and have a range of impacts, of most importance in the context of environ-mental sedimentology is the chemical remobil-ization of nutrients and contaminants duringearly diagenesis, and the release of contaminantsfrom floodplains.

1.5.1 Early diagenesis in aquatic sediments

Upon the consumption of O2, a series of anaerobicbacterial reactions are favoured, utilizing oxygenin species such as nitrate (NO3

2−), iron oxide(FeOOH), manganese oxide (MnO2) and sulphate(SO4

2−). These anaerobic early diagenetic reac-tions are many and complex. The most significant

Process Characteristics Deposit

Rockfall

Creep

Slide

Slump

Debris flow

Grain flow(fluidized flow)

Turbidity current D

ecre

ase

in c

once

ntra

tion

and

an

incr

ease

in in

tern

al d

isag

greg

atio

n Avalanche deposit

Creep deposit

Slide

Slump

Debrite

Grain/fluidized flow deposit

Turbidite

Fig. 1.11 Processes and deposits associated with rock and sediment movement along a continuum ofdecreased concentration and increased internaldisaggregation. (Adapted from Stow 1986.)

ESC01 9/8/06 2:40 PM Page 14


reactions are nitrate reduction, Mn(IV) reduction,Fe(III) reduction, sulphate reduction and meth-anogenesis. All of these reactions break downorganic matter and, therefore, lead to an overalldecrease in organic matter content as sedimentsare buried. Many of these reactions can only utilize simple organic molecules, such as acetateand hydrogen, as the reductant. However, somebacterial communities, particularly iron-reducingbacteria, have been shown to possess the abilityto utilize complex organic molecules (Lovley & Anderson 2000). Therefore, such diageneticreactions may act to break down persistentorganic contaminants in aquatic sediments.Bacteria can also directly mediate the reductionof some contaminant metals, for example Cr, U,Se, Hg and Tc (e.g. Lovley 1993).

Early diagenetic reactions have an impact uponthe short- and long-term fate of contaminants insediments through two principal mechanisms:release of contaminants into sediment porewaters;and the uptake of contaminants into authigenicmineral precipitates. The oxidation of organicmatter and the reduction of iron and manganeseoxides result in the release of contaminants asso-ciated with these mineral phases to sedimentporewaters (Rae & Allen 1993). These increasedporewater contaminant concentrations can resultin the molecular diffusion of contaminants intothe overlying water column (commonly termeda ‘benthic flux’). There is a growing awarenessthat benthic contaminant fluxes to intertidalenvironments can be as significant as riverineinput and may act as a major long-term input of contamination into water bodies. Rivera-Duarte & Flegal (1997a; b) documented thatbenthic fluxes of Co and Zn from sediments inthe San Francisco Bay were of the same magni-tude as riverine inputs. Similarly, Shine et al.(1998) showed that the flux of Cd and Zn fromcoastal sediments in Massachusetts, USA was of a similar magnitude to that within the watercolumn itself.

In marine and brackish intertidal sediment-ary environments, sulphate reduction is a majorpathway for organic-matter oxidation and as aresult sulphide is released into porewaters. Sul-phide forms a highly stable complex with most

metals (Cooper & Morse 1998) and consequentlymetals released by Fe(III) and Mn(IV) reductionwill be precipitated out as sulphides. These pre-cipitates are predominantly in the form of ironmonosulphides, and metals may be adsorbed ontothe sulphide surfaces, or incorporated into thesulphide structure (Parkman et al. 1996). Earlydiagenetic metal sulphides have also been docu-mented in mining-impacted estuarine sediments,acting as a long-term sink for contaminants inthese sediments (Pirrie et al. 2000; see Chapter 7).In contrast to natural sediments, early diageneticmineral precipitates within contaminated sedi-ments can be varied and unique. For example,Pirrie et al. (2000) described the occurrence ofearly diagenetic simonkolleite (a Zn–Cl mineral)from metal-contaminated estuarine sediments.Early diagenetic minerals (e.g. vivianite – ironphosphate) can also be important in non-marinesediments (see Chapter 6).

1.5.2 Remobilization from floodplains

Contaminants may also be remobilized fromriver floodplains. Floodplains are sites of sedi-ment accumulation within river basins and,therefore, are classically considered to be con-taminant sinks, thereby preserving good tem-poral records of contaminant input (e.g. Smol2002; Chapter 3). These sinks of contaminants,however, can also become sources as a result ofpost-depositional processes, both chemical andphysical. For example, Hudson-Edwards et al.(1998) demonstrated that remobilization of Pb,Zn, Cd and Cu within overbank sediments ofthe River Tyne, England, occurred as a result of changes in water-table levels and the break-down of organic matter above the water table(see Chapter 3).

Contaminants stored on floodplains also maybe remobilized through physical erosion, andthis may take place long after the primary con-taminating activity (e.g. mining) has ceased. Forexample, Macklin (1992) showed that the prim-ary source of Pb and Zn to the contemporaryRiver Tyne, northern England, was remobilizedfloodplain alluvium originally deposited duringeighteenth and nineteenth century metal mining.

ESC01 9/8/06 2:40 PM Page 15


Contaminant remobilization is often triggeredby natural (climate) or anthropogenic (land use)changes that cause modifications, first in sedimentload and delivery, and eventually in erosion anddeposition. Macklin (1996) warned that flood-plain contaminant remobilization is increasingas a result of the hydrological changes associ-ated with global warming, and stressed that the long-term stability of contaminant metals withrespect to changes in physical (river bank andbed erosion, land drainage and development) andchemical conditions (redox and pH) is poorlyunderstood. The remobilization of contaminantsthrough the physical erosion of contaminatedsaltmarsh sediment can also be a significant sourceof contaminants to estuaries and coastal waters(see Chapter 7).

1.6 SEDIMENTARY RESPONSES TO ENVIRONMENTAL CHANGE

1.6.1 Sedimentary responses to naturaldisturbance events

Although daily or ongoing processes of fluvialor tidal flow and water/wind velocity influencebackground levels of sediment transport andaccumulation, the amount of sediment transportthat occurs during ‘normal’ conditions is rela-tively low. Most sediment transport and, as aresult, much of the morphological change thatoccurs within sedimentary environments takesplace during low-frequency but high-magnitudeevents. These may be associated with storms orhigh (seasonal) rainfall episodes, or with episodichigh-energy events such as cyclones and tsunami.At these times, sediment transport rates can dra-matically increase and hence a high proportionof annual sediment movement may occur over aperiod of only a few days. This is particularly thecase in many arid and semi-arid environments,which are characterized by highly ‘flashy’ dis-charge events and where short-lived but high-intensity rainfall events lead to very high-energyflows (see Chapter 5). High-magnitude dischargeevents also characterize many seasonally influ-enced fluvial systems. In the Burdekin Rivercatchment, North Queensland, tropical cyclones

dramatically increase discharge rates throughthe catchment (Fig. 1.12a), resulting in increasedsuspended (Fig. 1.12b) and bedload sedimenttransport (Amos et al. 2004). These high-energyevents also influence spatial variability in sedi-ment transport and storage within fluvial catch-ments. For example, in the Rajang River Deltaof Sarawak, eastern Malaysia, sediments arestored on the delta plain during the ‘dry’ season,but undergo rapid offshore transport during the ‘wet’ season when discharge rates increase(Staub et al. 2000).

In coastal environments, storm waves aredirectly responsible for extensive reworking ofunconsolidated sedimentary deposits and this ismanifested in changes in beach profiles and thebreaching of coastal barriers (see Chapter 8), andthe on- and offshore transport of sediment andrubble (see Chapter 9). In the tropics, high windspeeds and storm waves associated with cyclonescan lead to tree damage and mortality withinmangroves. This, in turn, can facilitate substratedestabilization and erosion of intertidal sedi-ment substrates. These events can thus result in significant localized ecological damage andoften marked short-term changes in patterns of nearshore sediment accumulation. However,cyclones and other high-energy episodic eventsare also important controls on the longer termdistribution and development of sedimentaryenvironments. On the Great Barrier Reef shelf,for example, cyclones generate northward flow-ing alongshore currents, which result not only insignificant along-shelf sediment transport, butalso a marked partitioning of sediment across theshelf (Larcombe & Carter 2004; see Chapter 9).

1.6.2 Anthropogenic modifications of sedimentsand sedimentary systems

Although seasonal and natural (i.e. storm-induced) changes in energy levels can lead tochanges in sediment dynamics and accumulationrates, anthropogenic activities also have poten-tial to modify rates of sediment input, sedimenttransport pathways and the composition of theaccumulating sedimentary materials. Clear linksbetween anthropogenic activity and sedimentary

ESC01 9/8/06 2:40 PM Page 16


system response occur, for example, in areaswhere construction or resource extraction activ-ities in the upstream sectors of catchments resultin downstream sediment starvation and/or ero-sion. At a global scale, the transport of sedimentthrough river systems represents a major path-way of sediment movement from upland ‘source’to marine ‘sink’. The transport of such sediment-ary material is, however, highly sensitive to a range of anthropogenic influences, includingreservoir construction, land-use change, soil andwater conservation activities, and sediment con-trol programmes (Walling & Fang 2003). Someof these activities lead to increased sediment loads

but others, and in particular reservoir construc-tion, lead to reduced sediment transport. Theimpacts vary between catchments, but in someregions reduced sediment supply has resulted inmarked changes in the behaviour and geomor-phology of fluvial systems (see Chapters 2 and3). In the Alpine region of Europe, for example,sediment deficits have been recorded in manyrivers over the past 30–40 years. Such reduc-tions have resulted from excessive gravel extrac-tion from rivers and the retention of sedimentbehind dams. The result, on many upland rivers,has been widespread erosion and entrenchment(Descroix & Gautier 2002).

12000

10000

8000

6000

4000

2000

0

Dis

char

ge (

m3

s−1)

1995 1996 1997 1998 1999 2000

TC Barry (Gulf of Carpenteria)

TC Gillian

TC ItaTC Justin

Ex-TC Sid (embedded in monsoon trough)

Low pressure system.Record rainfall in partsof Burdekin catchment

Monsoon low

TCs incoral Sea

Low pressuresystem along coast

Monsoon troughand TC Steve

(a)

(b) 12000

10000

8000

6000

4000

2000

Dis

char

ge (

m3

s−1 )

22 23 24 25 26 27 28 29 1 2 3 4 5 6 7 8 February March

900

800

700

600

500

400

300

200

100

Ave

rage

sus

pend

ed s

edim

ent

conc

entr

atio

n (m

g L−

1 )

DischargeSSC

Fig. 1.12 (a) Data from hydrographs recording fluvial discharge between 1995 and 2000 in the lower part of the Burdekin Rivercatchment. (b) Detail of discharge and suspended sediment concentrations (SSC) between February and March 2000. Note that SSC generally decreased through the discharge event and that, in this catchment, SSC levels appear to be supply limited. (Adapted from Amos et al. 2004.)

ESC01 9/8/06 2:40 PM Page 17


Similar links between sediment sources andsinks are evident along many coastlines wheresediment is supplied either from fluvial sourcesor from one area of coastline to another, oftenvia longshore transport. Reductions in sedimentsupply often lead to increased rates of erosionalong the coastal sectors that are deprived ofsediment. This occurs, for example, where sedi-ments are trapped behind dams located on therivers that feed the coastal sector. In California,reduced fluvial sediment supply due to damminghas led to increased rates of cliff erosion (seeChapter 8), and extensive subsidence and shore-line erosion of the Nile delta is attributed tosignificantly reduced sediment supply to thedelta front (see Chapter 7). In the northern Gulfof California fundamental changes in the sourcesand rates of sediment supply, and in the composi-tion of accumulating sediment, have also beendirectly linked to the effects of dam constructionand have resulted in a 95% reduction in sedi-ment supply from the Colorado River (see CaseStudy 1.1). Coastal retreat may also occur where‘upstream’ sediment supply has been restrictedby sea-defence construction. In such cases, sea-wall or groyne systems either prevent or restrict

sediment throughput, leading to increased ratesof downstream beach erosion and shorelinechange (see Chapter 8). The recognition of theseimportant sedimentary links has been a keydriver in the development of integrated catch-ment and coastal management schemes.

Anthropogenic-related modifications to sedi-ment source and transport pathways may besignificantly exacerbated by the effects of urban-ization. In relation to sedimentary systems, themost important influence occurs where construc-tional activities occur within the sedimentologic-ally active zone. Problems arise either because ofconstruction in areas where episodic sediment-ological and geomorphological changes can beexpected, or where deliberate constructional act-ivities have a consequent effect upon pathwaysof sediment transport and zones of sedimentaccumulation. Coastal dune and barrier islandsequences, for example, form part of active sedi-mentary systems that will respond to high-energystorm events. Hence, roads and houses built insuch zones become susceptible to storm damage(see Chapters 8 and 9). Similarly, constructionof seawalls or other landward constraints maylead to ‘coastal squeeze’ as landward migration

Case study 1.1 Anthropogenic modifications to sediment supply, northern Gulf of California

The Gulf of California is a narrow epicontinental sea, 1500 km long, that has formed from tectonic activity along the Californian coast (Case Fig. 1.1). The Northern Gulf of California(NGC) receives sediment from four areas, the Colorado River, the batholith of the Baja California,the Sierra Madre Occidental, and the deserts of north-west Mexico. Each of these source areashas a distinctive mineralogical signature enabling the provenance of accumulating marine sedi-ments in the NGC to be determined. On this basis four distinct sediment provinces have beenidentified, these being, (i) the Colorado River Delta Province, (ii) the Concepción River Province,(iii) the Transitional Province and (iv) the Baja-Sonora Province (Case Fig. 1.1).

Historically, the Colorado River has been the primary source of sediment into the NGC, withan estimated annual sediment discharge of 160 × 106 t. Fluvial sediment supply has, however,been dramatically reduced over the past 100 years following the construction of a series ofdams along the river. Of particular significance have been the Hoover Dam (built in 1934) andthe Glenn Canyon Dam (built in 1952). The result of this extensive water flow regulation hasbeen to reduce fluvial sediment supply by around 95%, resulting in sediment starvation to bothestuarine and deltaic environments of the Colorado River mouth, and in the northern areas ofthe Gulf.

ESC01 9/8/06 2:40 PM Page 18


In the vicinity of the Colorado River,major changes in sediment supply andtransport have been identified and, as aresult, oceanic (rather than fluvial) hydro-dynamic forces now exert the major influ-ence on sediment dynamics within the estuary and delta (Carriquiry & Sánchez1999). Rather than a predominant north to south (fluvial to basinal) transfer of sedi-ment, sediment is now transported fromsouth-east to north-west along the easternside of the NGC into the estuarine basin,and then reworked southwards along thewestern sides of the NGC (Case Fig. 1.1).Despite significant reductions in fluvial sediment input from the Colorado River,however, average sedimentation rates inthis NGC are reported to have remained relatively constant over the past 100 years.This is attributed to a transition in thesource areas that supply sediment to theGulf (Carriquiry et al. 2001). In particular,a high proportion of sediment is now supplied from resuspension and reworking of the Colorado Delta sediments and from the shallower part of the NGC shelf. Thesesediments form the Colorado River DeltaProvince and dominate the central areas of the NGC marine basin (Case Fig. 1.1).Additional sources of sediment are derivedfrom the desert areas of north-west Mexicoand south-west USA. These form the Concepción River Province and feed into the basin via the Sonoita and Concepción Rivers (Case Fig. 1.1). In addition, intense desert winds from the Sonora Desert represent an important transport medium for aeolian sediment transportinto the Gulf. These sediments are rich in zircon and garnet, and contribute primarily to theTransitional Sediment Province. The area therefore emphasizes the effects of anthropogenic-ally influenced reductions in sediment supply through fluvial systems, and the consequent‘downstream’ impacts on both sediment transport pathways and on the composition of theaccumulating marine sediments.

Relevant reading

Carriquiry, J.D. & Sánchez, A. (1999) Sedimentation in the Colorado River delta and Upper Gulf of Californiaafter nearly a century of discharge loss. Marine Geology 158, 125–45.

Carriquiry, J.D., Sánchez, A. & Camacho-Ibar, V.F. (2001) Sedimentation in the northern Gulf of Californiaafter cessation of the Colorado River discharge. Sedimentary Geology 144, 37–62.

Sonoita River

ConcepcionRiver

Colorado River

PacificOcean

BAJACALIFORNIA

N

20 km

Colorado River Province

Transitional Province

Baja California Province

Concepcion Province

Prevailing sediment transport direction

Tidally-influenced sediment transport

USA

MEXICO

ColoradoRiver

Northern Gulfof California

Northern Gulf

ofCalifornia

Case Fig. 1.1 Distribution of the main sediment provinces inthe Northern Gulf of California. The main sediment transportpathways in the vicinity of the Colorado River Delta are alsoshown. (Adapted from Carriquiry & Sánchez 1999; Carriquiry et al. 2001.)

ESC01 9/8/06 2:40 PM Page 19


of coastal sedimentary environments is restricted.Such interactions with sedimentary systems or arestriction in the way sediment systems respondto increased energy levels often brings with it a management or remediation ‘cost’. Hence inmany cases, the need for management is oftendriven not so much by the actual event, but as a result of (increasing) human occupation andmodification of the environment, i.e. urbaniza-tion of environments that will naturally respondto changes in the energy inputs associated withstorm or flood events. The influence of urban-ization of the coastal fringe is seen particularlyclearly in relation to estuarine environmentswhere large areas of intertidal land have beenclaimed over a period of several centuries (seeChapter 7). The result is often a fundamentalchange in the character and extent of intertidalland, and a suppression of an estuary’s ability torespond to changes in nearshore energy regimesor sea-level state.

Urbanization also has major impacts uponthe hydrology of catchments and river basins,which in turn influences the nature of sedimentmovement and accumulation (see Chapter 6).The increase in runoff rate in urban systemsleads to enhanced flooding pressures in riversystems, and this is often exacerbated by thepast removal of floodplains and river culverting,which inhibits the accumulation of sediment.These increases in runoff rate also have markedimpacts upon sediment transport in urbanizedriver basins, with large storm events accountingfor the majority of suspended sediment trans-port flux over short periods of time (Goodwin et al. 2003; Old et al. 2003).

Another highly significant anthropogenicimpact on sediments is that of sediment com-position and quality. The increase in contamin-ant loading in sediments has been extensivelydocumented for virtually all sedimentary sys-tems globally, including those that are generallyassumed to be pristine. These have been docu-mented both through monitoring programmeson sediment composition, and sedimentaryarchives of contaminant accumulation, such assalt marshes, lakes, reservoirs and floodplains.The former approach allows for short-term data

sets only, as monitoring programmes on sedimentcomposition have not been in place for long.Longer temporal records of sediment composi-tion may be recorded by sediments accumulatingin depositional environments (see Case Study 1.2).The nature and length of the sediment recordwill depend on a number of factors, includingsediment accumulation rates, extent of sedi-ment disturbance and post-depositional changes.Examples of temporal records of sediment com-positional change for lakes and river basins canbe found in Chapter 4 and in Smol (2002). Suchcompositional changes, and associated recordsof environmental pollution have also beenclearly documented for saltmarsh sediments (see Chapter 7).

1.6.3 Response of sedimentary systems to climaticand environmental change

Given the influence that climate exerts on thedevelopment of sedimentary environments, ongo-ing and projected climatic and environmentalchanges are potentially significant in relation to the dynamics and functioning of most sedi-ment systems. Climate exerts, for example, animportant influence on weathering regimes, the hydrological cycle (including seasonality of rainfall) and the frequency and magnitude of high-energy (storm) events, all of which areimportant in determining rates of sediment supply and transport (e.g. Chapters 2–4). In themarine environment, climatic conditions alsoinfluence environmental factors such as levels of dissolved CO2 and sea-surface temperatures.These are primary controls on the distributionand development of biogenic sedimentary deposits(Chapter 9). Sea-level itself is a major controlboth on the distribution and extent of coastalsedimentary environments (Chapters 7 and 8)and, because it influences base level, a majorforcing factor with fluvial systems (Chapter 3).Hence many of the predicted changes in globalclimatic conditions need brief considerationhere. These include changes in atmospheric CO2concentrations, increased atmospheric and sea-surface temperatures, increased UV radiation,changes to patterns of storm frequency and

ESC01 9/8/06 2:40 PM Page 20


Case study 1.2 Temporal changes in sediment composition as a result of pollution: archives of lead pollution

Sediments that have accumulated in aquatic systems may be good records of the external inputsaffecting sediment composition, and can be used as both indicators of pollution impacts uponsediment composition, and as archives of local and global pollution changes. Extensive studieson lake, reservoir and river basins have been undertaken. This example focuses on lead in lakeand reservoir sediments. Lead (Pb) is a natural element, which is supplied to sediments throughgeological weathering of rocks and mineral deposits. However, Pb has also been sourced to sediments through two anthropogenic processes: the mining and smelting of Pb ores, and thecombustion of fossil fuels, especially vehicle fuel with added Pb. Levels of Pb in the environ-ment are a concern because Pb can act as a powerful neurotoxin.

Studies of sediment composition in lakes in northern Scandanavia have documented long-term(over 2000 yr) records of sediment-Pb composition (Brännvall et al. 1999). Sediment composi-tion in these studies (Case Fig. 1.2) shows consistent changes in Pb deposition to these sedimentsover a wide area. An early peak at around 2000 yr BP was related to lead smelting during theGreek and Roman cultures. This was followed by peaks in Pb deposition around ad 1200 and1530, again related to lead smelting and coinciding with known peaks in metal production in Europe. The presence of significant Pb in sediments of these ages indicates that sedimentcomposition was being markedly altered prior to the Industrial Revolution, and that pollutionlevels in the environment are not all a recent phenomenon. These lake sediments also show aclear input of Pb in the latter half of the twentieth century (Case Fig. 1.2), related to the burningof fossil fuels and the use of Pb in petrol. This late twentieth century impact upon sediments hasalso been documented by numerous other studies from a wide range of sedimentary environ-ments (e.g. Renberg et al. 1994; Shotyk et al. 1998; and see Chapter 6). The use of stable isotopesof Pb (206Pb/207Pb) further clarifies the increases of anthropogenic Pb pollution from natural Pbinputs from soil weathering (Case Fig. 1.2a).

Shorter time periods of Pb-pollution impacts on sediment composition have been studied in reservoir sediments, as these systems commonly display faster sediment accumulation rates,thereby allowing increased temporal resolution. For example, Callender & Van Metre (1997)studied sediment cores from water reservoirs in the south-west and Midwest USA (Case Fig. 1.2b).This study clearly documented a high Pb peak in sediment between 1970 and 1980, linked toatmospheric Pb input from leaded-fuel combustion. More recently deposited sediments displaya clear, and rapid, reduction in Pb levels, as a result of the Clean Air Act of 1970 in the USA,and also the phasing out of leaded-fuel.

Similar records of Pb pollution have also been documented from Scottish freshwater lakes(Eades et al. 2002). Sediments in Loch Lomond show an increase in Pb past 1800 as a result of industry and fossil-fuel burning (Case Fig. 1.2c). A large increase in the latter half of thetwentieth century, coupled with a significant drop in the 206Pb/207Pb of this Pb was a result ofvehicle combustion of fuel with Pb additives. A significant fall in Pb levels since the 1980s was a result of the increasing use of unleaded petrol in vehicles.

Relevant reading

Brännvall, M.-L., Bindler, R., Emeteryd, O., et al. (1997) Stable isotope and concentration records of atmosphericlead pollution in peat and lake sediments in Sweden. Water, Air and Soil Pollution 100, 243–52.

ESC01 9/8/06 2:40 PM Page 21


Brännvall, M.-L., Bindler, R., Renberg, I., et al. (1999) The Medieval metal industry was the cradle of modernlarge-scale atmospheric lead pollution in Northern Europe. Environmental Science and Technology 33,4391–5.

Callender, E. & Van Metre, P.C. (1997) Reservoir sediment cores show U.S. lead declines. EnvironmentalScience and Technology 31, 424A–8A.

Eades, L.J., Farmer, J.G., MacKenzie, A.B., et al. (2002) Stable lead isotopic characterisation of the historicalrecord of environmental lead contamination in dated freshwater lake sediment cores from northern and central Scotland. The Science of the Total Environment 292, 55–67.

Renberg, I., Wik Persson, M. & Emeteryd, O. (1994) Pre-industrial atmospheric lead contamination detected inSwedish lake sediments. Nature 368, 323–6.

Shotyk, W., Weiss, D., Appleby, P.G., et al. (1998) History of atmospheric lead deposition since 12,370 14C yrBP from a peat bog, Jura Mountains, Switzerland. Science 281, 1635–40.

1.2

1.3

1.4

1.5

40

30

20

10

0

1900 2000 AD700 900 1100 1300 1500 170050001000BC

Pollution Pb

206Pb/207Pb

Calendar Years

Pol

lutio

n P

b20

6 Pb/

207 P

b

1900

1920

1940

1960

1980

2000

0 20 40 60 80 100 120 140 160

Pb concentration (μg/g)

Age

White Rock LakeLake AnneLake Harding

40

20

0

1600 1700 1800 1900 2000

1.14

1.18

1.22Pb concentration206Pb/207Pb

Pb

conc

entr

atio

n (

μg/g

)

206 P

b/20

7 Pb

Date

a

c

b

Case Fig. 1.2 (a) Lead pollution and lead isotope profile preserved in an annually laminated lake, Lake Koltjärn, Sweden. (After Brännvall et al. 1999.) (b) Historical records of lead inputs preserved in urban and suburban lakes in Georgia, USA. (After Smol 2002; Callender & Van Metre 1997.) (c) Records of lead pollution and lead isotope signature in a Scottish freshwaterlake, Loch Lomond. (After Eades et al. 2002.)

ESC01 9/8/06 2:40 PM Page 22


intensity, and increased sea-levels. Dependingupon the magnitude of these changes, both posi-tive and negative responses may occur withindifferent sedimentary environments.

The long-term links between climate and sediment system response are clearly illustratedin studies that have examined the response oflarge fluvial systems to Quaternary or Holoceneclimatic change. The Ganges river system, forexample, represents one of the largest sedimentdispersal systems in the world. It extends for adistance of over 3000 km, presently dischargingaround 300 × 109 m3 of water and 520 × 106 t ofsediment annually (Goodbred 2003). Despite itsvast size, strong sediment linkages occur betweenthe source areas in the Himalaya, through thecatchment basins on the Ganges Plain, to thedepocentres in the Bengal Basin. This is believedto reflect the strong seasonal control that existson river flow, with around 80% of fluvial dis-charge and 95% of sediment load delivered overthe 4 month summer monsoon period. As aresult, the system is highly susceptible to changesin atmospheric circulation patterns and, in particular, any change in the strength of thesummer monsoon. This has altered several timesover the past 150 kyr, resulting in changes inrates and patterns of fluvial sediment produc-tion, erosion, transport and accumulation, andis manifested by shifts from periods of upstreamerosion and entrenchment, to periods of sedimentaccumulation on the delta plain and within theBengal Basin.

Large-scale studies thus establish clear linksbetween sediment dispersal system behaviourand climatic regime. Predicting future changesin sediment system dynamics is, however, com-plicated by the uncertainty that exists in rela-tion to the magnitude and regional variability offuture climate change. Future climate projectionsare based, in part, upon constructed scenarios offuture human behaviour (in relation to issuessuch as greenhouse gas emissions and land-usepractice). They are also dependent, however,upon atmospheric circulation patterns andatmospheric interactions with large-scale oceancurrents and land features (including albedo,vegetation cover and soil moisture content).

Among the more complex models in use are thecoupled atmosphere–ocean general circulationmodels (AOGCMs). Examples of these includethe recent HadCM3 model developed at theUK’s Hadley Centre (see Hadley Centre 2004),the projections from which are discussed below.Such models assume that future greenhouse gas emissions will follow the widely used IS92ascenario (Houghton et al. 1992), whereby CO2levels will double during the twenty-first cen-tury, there will be mid-range economic growth,and no measures made to reduce greenhouse gasemissions. The dynamics of, and interactionsbetween, the complex variables involved in suchmodels means, however, that although climateprojection models are progressively improving,there remain many unknowns and significantuncertainties about the feedbacks that occurbetween different parameters. Hence, a range ofclimate simulation models have been developedthat are based on different input data and accountfor different future scenarios (Hadley Centre2004; McCarthy et al. 2001).

Although future changes in variables such astemperature, rainfall and sea-level are clearlysignificant from a sedimentological perspective(outlined below), it is important that these projections are placed in the context of pastenvironmental change. The Quaternary has, for example, been characterized by marked climatic shifts associated with glacial and inter-glacial phases. At the global scale, these oscilla-tions resulted in marked changes in climate, but the actual effects were spatially very variable.Hence, although previously temperate regionsin the northerly latitudes cooled and were sub-ject to glacial or periglacial conditions, more aridregions to the south became more temperate incharacter. Consequently the nature of environ-mental change and, as a result, the processesinfluencing sediment transport and accumula-tion differed between regions. Such spatial vari-ability is likely to be a key factor in the contextof any future climatic change. It is also relevantto note that the magnitudes of projected futurechange in, for example, temperature and sea-levelare both above and below those experienced inthe recent past. During the early Holocene, for

ESC01 9/8/06 2:40 PM Page 23


example, global temperatures were rising at ratesof around 1–2°C/1000 years. This is below therates of most projections through to 2100 (seebelow). By contrast, sea-level has risen by around120 m over the last 17,000 years, and most ofthis rise occurred prior to 6000 ybp. Hence ratesof rise were in the region of 10 mm yr−1, con-siderably above current projections of change(Jones 1993a).

Although future projections are subject to ahigh degree of uncertainty, what is evident fromrecent data is that marked increases in atmo-spheric concentrations of greenhouse gases haveoccurred over the past 200 years. The followingincreases have, for example, been reported for theperiod 1750–2000: CO2 (280 ppm to 368 ppm),CH4 (700 ppb to 1750 ppb) and N2O (270 ppbto 316 ppb) (McCarthy et al. 2001). These areall predicted to increase further through to 2100(e.g. atmospheric CO2 concentrations are pre-dicted to increase to between 540 and 970 ppm).Linked to these greenhouse gas increases are pro-jected changes in global mean surface temper-atures, which will have potential consequencesfor sea-levels, sea-surface temperatures and clim-ate circulation systems. Global mean surfacetemperatures have, for example, increased by0.6 ± 0.2°C over the twentieth century, althoughthese increases have been more significant overland areas than the oceans. There has also beenan increase in the number of hot days and areduction in the number of days experiencingfrosts. In addition, Arctic sea-ice has thinned by around 40% in the past few decades anddecreased in extent by 10–15% since the 1950s(McCarthy et al. 2001).

Based on outputs from the HadCM3 models,which show differences between current climate(defined as the period 1960–1990) and climate atthe end of the twenty-first century (2070–2100),mean surface air temperatures are predicted to increase by 0.3°C (range 0.6–9.2°C; seeHadley Centre 2004). Increases are evidentacross much of the globe with the exception ofthe southern Pacific Ocean. Annual precipitationrates are also projected to increase by an averageof 0.2 mm day−1 (range − 3.7–8.9 mm day−1; seeHadley Centre 2004). These projections are more

variable regionally, but show increases over thenorthern mid-latitudes, tropical Africa and south-east Asia. Decreases are predicted in Australia,central America and southern Africa.

Such changes in temperature and rainfall are likely to have an impact upon the function-ing of many terrestrial sedimentary systems. Aspatterns of river channel erosion and sedimenta-tion are influenced by streamflow over time and,especially, by flood frequency, any changes inthe hydrological cycle will significantly influencefluvial sediment transport and depositional processes (see Chapter 3). Changes in precipita-tion may result in modified drainage densities andeither higher or lower sediment yields depend-ing on the regional effect. In either case therewould be significant change in the downstreamdepositional environments. Lakes are alsohighly susceptible to changes in air temperatureand rainfall because these influence rates ofevaporation, lake-level and hydrochemical andhydrobiological regimes (see Chapter 4). Underextreme conditions, lakes may disappear entirely.Responses are also likely to vary between open(exorheic) and closed (endorheic) lakes. The latter are dependent upon rates of fluvial inputand evaporation and are thus highly sensitive tochanges in both. Hence lake sediments are goodsources of information about past climatic andenvironmental conditions.

Climate changes are also likely to have animpact upon rates of sea-level rise and on thefunctioning of large-scale ocean–atmospherecirculation systems. Over the past 100 years,global mean sea-level has increased at an aver-age rate of 1–2 mm yr−1. Predictions from theHadCM3 models suggest that mean sea-levelwill rise by 0.38 m by the end of the twenty-first century (range 0.09–0.74 m; see HadleyCentre 2004). This will occur primarily becauseof thermal expansion of the oceans and themelting of glaciers and ice caps. Significantregional variations in the magnitude of theserises are likely, however, superimposed uponwhich will be spatial variations in rates of isostatic change. Even relatively small increases in sea-level will, however, exert a significantinfluence on most coastal sedimentary systems

ESC01 9/8/06 2:40 PM Page 24


(e.g. Nicholls 2004) and increase rates of coastalcliff failure (Jones 1993b). In large part, this will occur because increasing sea-levels raise theplane of activity over which wave influence isexerted. Along temperate sediment-dominatedcoasts, this is likely to result in a landwardmigration of beaches, dunes and barrier islands,although the potential for landward retreat will depend upon the nature of the backshoreenvironment and on the presence or absence ofcoastal infrastructure. Where roads and urbanconurbations exist, landward migration may beprevented by coastal defence structures andthus, in such cases, progressive loss of coastalsedimentary environments (‘coastal squeeze’)may occur (Chapters 7 and 8). Rates of changewill be exacerbated along low-lying coasts andin areas that are undergoing active subsidence(see Pirazzoli 1996, fig. 117).

Predicting the response of individual coastalsystems is further complicated by local sedimentdynamics and rates of sediment supply. Coastaldune systems represent an interesting examplewhere these factors are highly site-specific andthus shoreline response modes highly variable.Dunes are sites of temporary sediment storageon the coast and significant sediment exchangeoccurs with adjacent environments. This may bea response to changes in energy levels duringstorm events, but will also occur during periodsof rising sea-level as the plane of wave influenceis reset (see Chapter 8). Hence dune responses tosea-level change, as with most coastal systems,can be considered realistically only on a site-by-site basis. In the tropics, sea-level rise may bringthe benefit of renewed phases of growth on coralreefs that have reached sea-level and ceased toaccrete during the late Holocene. The potentialfor such expansion will, however, depend uponreef community status and in severely degradedcoral reefs partial submergence of the reef struc-tures may occur. This, in turn, may facilitateincreased wave-overtopping and adjacent shore-line erosion (see Chapter 9). Along mangrove-dominated coasts, shoreline response to sea-levelrise will depend, in part, upon whether sedimentsupply is sufficient to maintain the seawardfringe. As along temperate coastlines, landward

migration of mangroves may occur if the back-shore area is unimpeded. Hence environmentalresponses are likely to be highly site-specific.Predicting future responses to sea-level rise alongmany coastal fringes is further complicated bythe fact that sea-level is rising from a position ofpre-existing sea-level highstand. Hence manycoastal sedimentary environments will be migrat-ing across areas of relatively low-lying land andthus the conditions under which the environ-ments transgress will be very different to thosethat were submerged during the Holocene sea-level rise. This will complicate attempts to modelfuture landform migration on the basis of changesthat occurred during the early Holocene.

Climate change models also predict an increasein the frequency of severe weather conditions(McCarthy et al. 2001). These include not onlystorms and cyclones, but also shifts in largescale, ocean-driven climate oscillations such asthe El Niño–Southern Oscillation (ENSO) andthe North Atlantic Oscillation (NAO). There is, for example, some evidence to suggest that El Niño events have become more frequent andintense during the past 20–30 years and thatthese events may increase in frequency and inten-sity through to 2100 in tropical Pacific areas.Such changes may result in modified rainfallpatterns, in rainfall intensity and the frequencyof drought conditions. Predicted increases inhigh-intensity precipitation events may result inincreased flooding, landslides and mudslides,increased rates of soil erosion and increasedflood runoff, with obvious implications both forupland and fluvial sedimentary systems. Pastrecords of increased slope failures in Central andSouth America have, for example, been linkedto interannual variations in the magnitude andintensity of rainfall events between El Niño and La Niña years (Trauth et al. 2000). Anychange in storm frequency and intensity willalso have significant implications for manycoastal sedimentary systems due, in large part,to likely increases in storm-wave surges. Underthese conditions, increased rates of coastal ero-sion (Chapter 8) and, in the tropics, increaseddamage to coral reefs and mangroves can beexpected (Chapter 9).

ESC01 9/8/06 2:40 PM Page 25


1.7 ASSESSMENT AND MANAGEMENT OF SEDIMENTS

Sediments are managed on many scales (bothspatially and temporally) for a wide range ofreasons, and different environments have uniquemanagement challenges and solutions (see indi-vidual chapters). In many cases it can be recog-nized that there is a problem from the sedimentperspective which needs to be addressed. For

example, sediment may be having an adverseimpact upon the ecological functioning of a riverbasin, or excess sediment accumulation may behaving a negative impact upon the economicfunctioning of a port or other navigable water-way. In most cases, these issues will be abouteither sediment quantity or sediment quality,although in many cases they are interlinked (see Case Study 1.3).

Case study 1.3 Sediment dredging and treatment in the Port of Hamburg, Germany

Hamburg harbour is the largest port in Germany, and one of the ten largest in the world. It isnear the mouth of the River Elbe, approximately 100 km from the North Sea, which drains central Europe (Case Fig. 1.3A). As it is an economic port, water depths need to be maintainedto allow shipping access into the port. Sediment is supplied to the port both from the upstreamcatchment of the River Elbe, but also by tidal movement of sediment from the North Sea.Sedimentation rates in some parts of the port are in the order of several metres per year, and thissediment accumulation has a major impact upon shipping access. As a result, there is a need forsediment to be dredged from the port; approximately 3 to 4 million cubic metres of sedimentfrom the Elbe each year.

This creates an issue of how to dispose of this dredged sediment. Before the 1970s thisdredged material was either placed on agricultural land or disposed of further down the system.However, high levels of contamination in this sediment, together with stricter environmental

Germany

CzechRepublic

ELBE

Border ofcatchment area

N

50 km

Hamburg

Prague

Case Fig. 1.3A The drainage basin of the River Elbe, which drains central Europe and has near its mouth the harbour ofHamburg, the largest port in Germany.

ESC01 9/8/06 2:40 PM Page 26


controls, have meant that this dredged sedi-ment can no longer be disposed of in thismanner.

Typical levels of metal and organic con-taminants in the dredged sediment are shownin Case Table 1.3. The largest portions of the contaminants present in the sedimentsdeposited in the port have been sourced from discharges in the upper reaches of theRiver Elbe. The upper Elbe drains industrialregions in the Czech Republic, and it is theseregions that contribute the largest contamin-ant load. The levels of contaminant inputsinto the port have decreased in recent years,partly as a result of environmental legislationon sources, but also due to economic declineand decreases in industrial activity.

Given the problem of high contamination levels, the Port of Hamburg has developed adredged sediment treatment process, whereby dredged sediment is treated, cleaned and thewaste material minimized (Case Fig. 1.3B). Contaminants in the dredged sediment are con-centrated in the fine fraction (< 63 μm), with the sand-sized fraction (> 63 μm) having muchlower levels of contaminants. Therefore, this treatment separates these two size fractions. The

Dewateredclay and siltExcess water to

waste-water treatment

Rotaryscreen

Coarsefragments

Dewateredsand

Hydro-cyclones

Fluidizedbed

Sanddewatering

screen

Dredged materialfrom harbour

Lamellarthickener

Processwater

Agitatingtank

Screen beltpress

High pressurebelt press

Case Fig. 1.3B The stages in the treatment of contaminated dredged sediment taken from the harbour of Hamburg. (Adaptedfrom Kroning 1990.)

Case Table 1.3 Levels of contamination in sedimentdredged from the harbour of Hamburg. (From Netzband et al.2002.)

Contaminant Levels in Hamburg Port dredged material (μg g−1)

Arsenic 50–150Lead 150–300Cadmium 5–25Chromium 150–300Copper 250–600Nickel 50–100Mercury 5–20Zinc 1000–2500Mineral Oil Up to 3000PCBs Up to 1.5PAHs Up to 15

ESC01 9/8/06 2:40 PM Page 27


Sediment management approaches generallytake one of two forms.1 Those that address issues resulting from thepresence of sediment (either a quality or quantityissue) and where sediment requires removing orremediating. This type of management is mostcommon in engineered or contaminated envir-onments, e.g. urban environments (Chapter 6)and estuaries (Chapter 7).2 Those that use management strategies to trapor retain sediments in a system to maximize the ecological or environmental functioning ofthat system. Such management practices areparticularly common in coastal environments(see Chapters 7–9).

1.7.1 Risk assessment and sediment guidelines for contaminated sediments

A first step in addressing these issues is to assessthe environmental risk of the sediments, and take

appropriate action. In order to determine if sedi-ment is contaminated, baseline and threshold-effect information is needed. In the case of artificial compounds (e.g. pesticides, PCBs, someradionuclides) the baseline value is zero, andcontamination assessment is relatively straight-forward. Pre-industrial historical values need tobe established for elements with both natural andanthropogenic sources (e.g. metals, phosphorus).In rare cases this can be accomplished throughanalysis of monitoring data or archived samples,but in most cases values are determined throughthe analysis of sediments that have accumulatedthrough time (Smol 2002). Threshold-effect valuesare often determined through an assessment ofthe physical, chemical or biological nature of thesediment, or a combination of these. Increasingly,the end result is a series of threshold-effect valuesthat collectively form Sediment Quality Guide-lines, or other similar measures (e.g. critical level,critical load, etc.).

steps in the treatment process are shown in Case Fig. 1.3B. The first stage of the process is theremoval from the wet sediment of most of the coarse fraction by centrifugation. Any remain-ing fine grains with the sand fraction are then removed in a fluidized bed under flowing water conditions. Sand-sized material is dewatered and can be used in the construction industry. Thefine-grained material is dewatered through a series of flocculation and filtration procedures.The resulting contaminated fine sediment is disposed of in a purpose-built disposal facility.Annually, 1.2 to 1.4 million cubic metres of sediment are treated in this way, with 50% of thisvolume being contaminated fine sediment placed in the disposal facility.

The removal of contaminated sediments via this dredging, sediment treatment and disposalhas removed about 30% of the heavy metal contaminants from the River Elbe and a similarpercentage of the organic contaminants. Therefore, this dredging acts as a pollutant filter to theNorth Sea. This is not a sustainable solution, however, and it is now widely recognized thatminimization of contaminants at source (i.e. in the upper reaches of rivers) is the most effectivesediment pollution strategy.

Relevant reading

Adams, M.-S., Ballin, U., Gaumert, T., et al. (2001) Monitoring selected indicators of ecological change in theRiver Elbe since the fall of the Iron Curtain. Environmental Conservation 28, 333–44.

Kroning, H. (1990) Separation and dewatering of silt from the Port of Hamburg. Aufbereitungs-Technik 4,205–14.

Netzband, A., Reincke, H. & Bergemann, M. (2002) The River Elbe: a case study for the ecological and eco-nomical chain of sediments. Journal of Soils and Sediments 2, 112–16.

ESC01 9/8/06 2:40 PM Page 28


Unlike water quality, clear Sediment QualityGuidelines are generally lacking for sediments,although many countries have developed theirown guidelines independently. These guide-lines can take chemical, biological or integratedchemical and biological forms. Simple chemicalguidelines involve either analysing for a con-taminant and setting acceptable levels based onnatural background levels (e.g. Ingersoll et al.1996), or assessing the potential chemical bio-availability of contaminants in a sediment (e.g.Ankley et al. 1996). Although such guidelines aresimple and relatively cheap to implement, theyare not based on an ecological response and,therefore, do not provide realistic informationon the risk associated with that contaminantlevel. Much better information may be acquiredthrough determining the ecological response tothe sediment. The most appropriate way to do thisuses ecotoxicity tests. Such tests use organisms(either benthic or water column) to determine thetoxicity of a sediment, the benefit of these beingthat they provide information on ecological risk.However, the test does not provide informationon which contaminant (or combination of con-taminants) is responsible for this risk. Recently,a triad approach has been gaining favour(Sediment Quality Triad; Chapman 1986). Thisassessment consists of three components:1 identification and quantification of all con-taminants in a sediment;2 measurement of toxicity based on a sedimenttoxicity test;3 evaluation of in situ biological effects.The benefit of this measure is that it integrateslaboratory, field, biological and chemical data.The disadvantages are that it has not been fullyaccepted and is expensive.

1.7.2 Managing issues of sediment quantity and quality

Once a sediment has been determined to be aproblem, either as a result of quantity or qualityissues, there are a number of remediation stra-tegies that can be put in place to address this.These can be placed into three main categories:physical removal of material (dredging), phys-

ical isolation of the material (containment) andchemical treatment of the sediment. Dredging ofsediment can be carried out for reasons of qual-ity, but by far the most important reason is forquantity reasons: for example, the need to dredgesediment to maintain a minimum draft for ship-ping in docks, channels and canals (see CaseStudy 1.3). Once such sediment has been dredged,however, there is commonly a secondary qualityissue as the sediment may be too contaminatedto dispose of it in a normal manner. Physicalcontainment is used in cases of sediment qualityissues and involves covering the contaminatedsediment, either with clean sediment or concrete.This procedure effectively isolates the contamin-ated sediment from the overlying water columnand ecosystem, and may often be the cheapestalternative. Chemical treatment of sediment cantake place ex situ or in situ. Ex situ treatment isgenerally used to produce sediment of low enoughcontamination levels to be disposed of safely orreused (see Case Study 1.3). In situ treatment of sediments is a novel application and, in a similar manner to biochemical remediation ofcontaminated land, is often site-specific, depend-ing on the problem.

1.7.3 Managing for sediment retention andstabilization

A number of sediment management strategiesfocus on either retaining or trapping sedimentfor the purpose of limiting change within thesedimentary system. Most commonly this is donein response to either one-off erosional events,such as can occur following cyclone or tsunamidamage, or in response to the progressive sedi-ment loss that occurs when the volume of sedi-ment being supplied to a particular environmentis reduced, i.e. there is a shift to a negative sedi-ment budget. Such problems are particularlycommon in coastal environments and reflect aneed to protect coastal infrastructure. In manycases the need for shoreline protection occurs dueto inappropriate siting of properties or roads inareas that should be expected to be periodicallyinfluenced by marine processes. The techniquesused are varied, but include:

ESC01 9/8/06 2:40 PM Page 29


1 the construction of shore normal groynes thatfacilitate the trapping of sediment being movedalongshore;2 the artificial emplacement of sediments ontobeaches to maintain their morphological integrity– a process known as beach recharge;3 the artificial stabilization of coastal dunes inan attempt to retain sediment within the dunesystem – numerous approaches including theuse of fences to trap windblown sand and bio-degradable fabrics to stabilize sediments havebeen used (see Chapter 8).Many of these techniques are of short-termbenefit only and may bring with them either acontinual maintenance cost (e.g. in the case ofongoing beach recharge), or may have down-drift sedimentological consequences (e.g. in thecase of groyne construction).

1.8 FUTURE ISSUES AND RESEARCH

As discussed at the start of this chapter, envir-onmental sedimentology represents a rapidlyexpanding field of research. It is increasinglyclear that most active sediment systems are, tovarying extents, influenced by human activityand thus significant attention is being targetedat trying to understand the impacts and effectsof human activity on contemporary sedimentsystems. One of the aims of this introductorychapter has been to provide an overview of the nature of such influences. These include, at a generic level, the modification of sedimenttransfer and accumulation pathways, the con-tamination of actively accumulating sediments,and modifications to sediment chemistry and sedi-ment diagenesis. Examples of each are includedin individual chapters, but in many cases there isa need for much higher resolution monitoring ofsedimentological and geochemical processes inorder to enhance understanding of the temporaland spatial dynamics of individual sedimentarysystems. An excellent example of the advancesthat can be made by adopting more integratedsedimentological, geochemical and ecologicalapproaches, as well as airborne sensing and satel-lite imagery, is seen in the recent developments

that have been made in understanding the dy-namics of the Amazon–Guianas coast in SouthAmerica (see Baltzer et al. 2004 and referencestherein). Such bodies of work demonstrate thebenefits of small-scale, localized studies that are placed in the context of an improved under-standing of regional sediment dynamics. Inmany sedimentary environments there are alsocomplexities emerging in terms of distinguishingbetween natural (background) sediment dyna-mics and those associated with anthropogenicactivities. Addressing such issues, as well asunderstanding the interactions between naturaland anthropogenically induced change, representimportant areas for future research.

Considerable research attention is also currently being focused on issues of sedimentmanagement and remediation, both in relationto addressing issues of sediment pollution, aswell as sediment abundance. There is a growingawareness that a physical and chemical under-standing of sediments needs to integrate waterand ecological information to better informhabitat monitoring and management (e.g. seeChapter 10). It is also important that the fullrange of scales on which sedimentary processesoperate is considered in sediment management.For example, within river basins a number ofmodels have recently been produced to managesediment at the catchment scale (Apitz & White2003; Owens et al. 2004). Such approaches are important for compliance with, and imple-mentation of, legislations and guidelines; forexample, the European Community WaterFramework Directive (2000/60/EC). Sedimentis also increasingly being considered as an eco-nomic, ecological and social resource, and assuch needs to be managed sustainably.

Overriding many of these themes are thepotential changes that will be associated withclimate-change-induced shifts in, for example,atmospheric and oceanic temperatures, rainfallpatterns, sea-level, oceanic CO2 concentrationsand storm frequencies. Understanding or predict-ing the likely response of sediment systems (bothterrestrial and marine) to such changes is, in part,reliant upon improved climate models. There isa need, however, for much better understanding

ESC01 9/8/06 2:40 PM Page 30


of likely sediment system responses at a range ofspatial scales. Large-scale studies, such as thoseof Goodbred (2003) that are outlined in section1.6.3, are useful in terms of providing an insightinto the way in which climate can control large-scale sediment system evolution. However, evenmodelling more localized sediment responses is fraught with difficulties and is constrained bythe limitations that exist in terms of the datathat can be reliably placed within appropriatemodelling software.

In the immediate future the most pressingresearch needs are likely to be driven by the fact

that many human-induced as well as environ-mental or climatic-induced changes to sedimentsystems bring with them a ‘human cost’. This willincreasingly occur where change within any indi-vidual sediment system progressively impingesupon human usage of the environment. This mayoccur either because of increased or decreasedsediment accumulation, or because contamin-ant accumulation is having an impact on theecological and/or sedimentary functioning ofthe environment. Understanding, mitigating andmanaging these changes represent major researchchallenges for the near future.

ESC01 9/8/06 2:40 PM Page 31

2.1 INTRODUCTION AND BACKGROUND

2.1.1 Introduction

Mountain environments account for approx-imately one-third of the Earth’s surface and it is estimated that about 10% of the world’s population live in mountain and upland regions(Gerrard 1990). Geomorphologically these areamongst the most active areas on Earth, oftenbeing characterized by some of the highestrecorded rates of erosion and sedimentation(Walling & Webb 1983; Jansson 1988). Moun-tain regions generally have steep slopes andlarge relative relief. These are factors that makemountain environments sensitive to natural andanthropogenic activity such as extreme clim-ate events, seismic perturbations, deforestationand land-use change. The dynamic nature ofmountain environments, however, must not beoveremphasized at the risk of ignoring slow,continually acting processes, the cumulative effectof which can be highly significant (Messerli 1983).Furthermore, sedimentary activity in mountainenvironments varies enormously between differ-ent topographic settings (Milliman & Sivitski1992) and even within the same general setting;differences caused by variations in the pressuresposed both by natural and anthropogenic agentsproduce differing sedimentary responses (Dedkov& Moszherin 1992).

An understanding of the environmental sedimentology of mountain environments isimportant because mountains provide essentialresources such as water supply, sustainable energy(hydroelectric power), recreation and tourism,

2Mountain environments

Jeff Warburton

ecological refuge and specialist agriculturalniches. Their significance varies from country to country and is generally proportional to thedegree of mountain cover, for example in Europethe proportion of area greater than 1000 m ofaltitude varies markedly: 5% in the UK, 44% inNorway and 100% in Andorra. Mountain areas,in common with many other environments, are increasingly endangered by socio-economicchanges, increased recreation and traffic, andchanging land-use, often leading to environ-mental degradation. These changes often can berelated to direct impacts such as building anddevelopment works, or may be the result of sub-tle environmental changes such as changing pre-cipitation patterns, shifts in habitats, changes inrunoff rates, water and soil pollution and changesin the ground thermal regime. Environmentaldegradation of this type is often manifest inchanges in sedimentary processes acting in moun-tain areas and may result in slope instabilitiesand enhanced fluvial erosion and sedimentation.An understanding of the environmental sedi-mentology of mountain areas provides a usefulframework for studying the effects of humansand environmental change on active surface sedimentary systems. Such an approach is par-ticularly pertinent to mountain environmentswhere, due to high relief and steeper slopes, geo-morphological and sedimentary processes oftenoperate at greater rates than in lowland environ-ments and the extreme physical conditions makemountain environments susceptible to even slightchanges in climate and land-use.

Given the large range in mountain environ-ments it is impossible in this short chapter to fully

ESC02 9/8/06 11:24 AM Page 32

MOUNTAIN ENVIRONMENTS 33

characterize the environmental sedimentologyof all mountain areas. The aim is, therefore, tohighlight the general functioning of active sedi-mentary processes in mountain environmentsand, using case studies and examples, show howthese processes are affected by climate changeand human actions and how adjustments to sed-iment systems have an impact on human use ofmountain regions. The chapter begins by con-sidering the main characteristics of mountainenvironments (section 2.1.2) that are most relev-ant to environmental sedimentology. This formsa basis for classifying mountain environments(section 2.1.3). The bulk of the chapter that follows is divided into five main sections: sedi-ment sources and transfer processes (section 2.2);processes and impacts – events related to naturaldisturbance (section 2.3); processes and impacts– anthropogenic influences (section 2.4); manage-ment and remediation (section 2.5); and futureissues (section 2.6). Inevitably some topics, suchas climate change, depositional processes (i.e.alluvial-fan and glacial-lake sedimentation) andsome slow geomorphological processes (e.g.creep), are not covered in great depth.

2.1.2 Characteristics of mountain environmentsrelevant to environmental sedimentology

There are seven general themes that need to beconsidered when dealing with the environmentalsedimentology of mountainous areas.1 Mountains are generally regions of abundantsediment supply and high erosion potential.Typically erosion rates and sediment yields areglobally some of the highest recorded (Milliman& Syvitski 1992).2 High rates of sediment production translateinto elevated rates of sediment transfer andincreased sediment deposition. However, sedi-ment delivery in tectonically stable and tectonic-ally active regions differs markedly dependingon the nature and rate of the different geomor-phological processes that operate (Marutani et al. 2001).3 The importance of steep slopes is fundamentalto many processes operating in mountain envir-

onments. High levels of gravitational stress resultin rapid sediment movements both in terms ofslope instability and channel sediment transport(Jones 1992).4 There is considerable variability in the spatialand temporal rates of sediment transfer and thishas important environmental and social con-sequences for mountain environment develop-ment (Butler et al. 2003).5 Mountain environments are sensitive to dis-turbance both from climate change and anthro-pogenic impacts (Ives & Messerli 1989).6 The incidence of geomorphological hazardstends to be high in mountainous, high-energyenvironments where narrow valley floors arejuxtaposed with steep unstable side slopes.Infrastructure and population are always at risk and this risk increases as expansion of roads and settlements continue. These issues are often greatest in mountainous terrain wherepopulation and infrastructure have developedalong upland river valleys. There is, however,growing recognition that heavy engineeringapproaches designed to manage such active geomorphological settings are unsustainable. Agreater understanding of catchment-wide sedi-ment budget dynamics can provide the neces-sary knowledge to better manage such systems(Gerrard 1990).7 Mountain sediment systems are often only apart of a larger drainage basin structure. There-fore, sediment delivery from the headwaters will have an impact downstream on floodplainprocesses. The degree of coupling needs to beestablished so that floodplains can be managedeffectively (Brizga & Finlayson 1994; Piégay et al. 2004) and hazards at the mountain frontreduced (White et al. 1997).

2.1.3 Definition and classification of mountain environments

Mountains occur in virtually all geographicaland climate settings on Earth. Mountain areasvary significantly from small isolated moun-tains to huge mountain chains stretching formany hundreds of kilometres across continents.

ESC02 9/8/06 11:24 AM Page 33

34 JEFF WARBURTON

(a)

(b)

Fig. 2.1 Two contrasting mountain types: (a) typical hilly terrain of Dooncarton Mountain in County Mayo,Ireland, 260 m a.s.l.; (b) high-mountain relief of Hunza Peaks in the Karakoram, Pakistan, 7000 m a.s.l.(Photographs: J. Warburton.)

ESC02 9/8/06 11:24 AM Page 34


Classifying different mountain systems is prob-lematic and several schemes have been pro-posed to account for the variability (Price 1981;Gerrard 1990). As a result, the definition ofmountain regions is largely arbitrary becausemultiple criteria can be used to define such areas,for example relative relief, the presence of par-ticular geomorphological features, a thresholdaltitude (1000 m), etc. Whatever definition isused, however, mountain environments haveseveral common features, namely: the presenceof steep slopes, and vertical differentiation ofclimate and vegetation cover (Barry 1992). Invery general terms as altitude increases relief is greater, vegetation cover diminishes and theclimate becomes more extreme in terms of precipitation, wind and temperature. Thesecharacteristic elements (greater relief, diminishedvegetation, extreme climate) are highly signific-ant in terms of the environmental sedimentologyof mountain regions because of their potentialimpact on erosion, i.e. the climatic control onweathering and therefore sediment production,the high energy associated with steep slopes andthe transport and removal of sediment, and thediminished vegetation which decreases resist-ance to erosion. In this chapter mountains aredefined as areas of steep relative relief wheresedimentation and erosion are actively condi-tioned by hillslope and/or channel processes. Thisgeomorphological definition avoids the problemof specifying a minimum altitude for moun-tain relief and can include contrasting examplesof the type shown in Fig. 2.1: (a) DooncartonMountain in Co. Mayo, Ireland (260 m of relief,maritime temperate climate) and (b) the HunzaPeaks in the Karakoram, Pakistan (3000 m ofvertical relief, arid continental interior). Processes

operating in the two environments are markedlydifferent.

Globally, mountain environments cover allclimate regimes and, as seen from the exam-ples just quoted, can range from small isolatedcoastal peaks to immense mountain ranges containing the world’s highest summits. Louis(1975), as part of an assessment of global relief,provides one of the few estimates of the area of mountain and high plateaux (Table 2.1).Assuming the total land surface is approxim-ately 149 million km2 (oceanic islands cover 2 million km2), mountains occupy about 20%of this area. The distinction between mountainand plateau relief is somewhat arbitrary so the total extent of world mountain area is open to some debate. However, the main pointis that mountains form a significant proportion(approximately one-fifth) of the global landarea. This global estimate is interesting but interms of mountain sediment systems it is the different relief regimes that are most importantin determining the processes operating within aparticular environment. Barsch & Caine (1984)distinguish four categories of mountain relief(Table 2.2), varying from subdued hilly terrainto high mountain systems (e.g. Fig. 2.1). Figure2.2 shows maps of (a) the major relief elementsof the Earth surface (Dongus 1980; as cited inBarsch & Caine 1984) and (b) the global dis-tribution of suspended sediment yield (Walling& Webb 1983). Although some exceptionsexist, such as the low suspended sediment yieldsfrom areas of old tablelands and continentalshields, the correspondence between areas ofhigh relief and active tectonics and seismicity,and elevated suspended sediment delivery at theglobal scale is clearly apparent.

Table 2.1 Global estimates of areas of mountain and plateaux. (Source: after Louis 1975; Barry 1992.)

Altitude range (m) Mountains (106 km2) Plateaux (106 km2) Mountain cover as a proportion of total land surface (%)

> 3000 6 4.02000–3000 4 6 2.71000–2000 5 19 3.4

0–1000 15 92 10.2Total 30 117 20.3

ESC02 9/8/06 11:24 AM Page 35

36 JEFF WARBURTON

Research themes in mountain geomorphologyhave been discussed by Barsch & Caine (1984),who make the important distinction (Fig. 2.3)between studies focusing on mountain landforms

and those that study land-forming processes inmountain environments. In terms of research intothe environmental sedimentology of mountainsystems it is the morphodynamics of mountains

Continental shields

Table lands

Cuestas, eroded old mountain systemsMountain - high mountain systems

Alluvial plains

Ice shields

> 1000

750–1000

500–750

100–250

50–100

0–50

250–500

Deserts and permanent ice

(a)

(b)

Fig. 2.2 (a) Global map of the major relief elements of the earth surface (Dongus 1980; as cited in Barsch & Caine 1984). (Reproducedwith permission, Mountain Research and Development.) (b) Global distribution of suspended sediment yield. (From Walling & Webb1983, Background to Palaeohydrology. (Ed.) Gregory, K.J. (1983). © John Wiley & Sons Limited. Reproduced with permission.)

Type Altitudinal difference (m) Relative relief (m km−2)(over 5 km distance)

High mountain system > 1000 500Mountain system 500–1000 200Mountainous terrain 100–500 100Hilly terrain 50–100 50

Table 2.2 Contrasting relief regimesfrom different mountain systems.(Source: Barsch & Caine (1984).Reproduced with permission MountainResearch and Development.)

ESC02 9/8/06 11:24 AM Page 36


that are of greatest interest and in particular the process dynamics and activity, because thisis what is most closely related to sedimentary processes. Process dynamics, however, can neverbe viewed in isolation because these are closelycontrolled by morphoclimatic factors as well as morphometry, structure and relief. Environ-mental sedimentology, therefore, embodies thefull range of research themes in mountain geo-morphology and in the case studies that followeach of these themes is covered.

2.2 SEDIMENT SOURCES AND TRANSFER PROCESSES

2.2.1 The mountain sediment cascade

An important conceptual model in understand-ing the environmental sedimentology of moun-tain environments is the notion of the mountainsediment cascade (Caine 1974). In simple termsthis depicts the mountain sediment system as a series of sediment stores linked by a series oftransfer processes (Fig. 2.4). Recognizing andquantifying significant sediment stores and theprocesses that link them is the basis of the sediment budget approach. The seminal workby Rapp (1960) working in the Kärkevaggecatchment in Sweden used this methodologicalapproach and inspired others to follow a similarapproach in identifying the significance of sedi-ment storage in regulating the sediment yield

from mountain catchments (Church & Ryder1972). Caine (1974) provided an early over-view of alpine geomorphological processes and described these as a series of sediment flux cascades. This consisted of two dynamic geomor-phological subsystems, namely the slopes andstream channels (Fig. 2.4). Caine (1984) subse-quently developed this concept by superimpos-ing three sediment subsystems: the geochemical,the fine sediment and the coarse detritus. Thisbasic description was later modified by Barsch& Caine (1984) into a four-way classification ofmountain processes each with different controls,responses and rates of activity.

Barsch & Caine (1984) distinguished:• valley glacier sediment system• coarse debris system• fine sediment system• geochemical systemAll four systems interact and material is trans-ferred between the different systems and henceit is often convenient to couple some of thesesystems together. For example, the valley glacierand coarse sediment subsystems are most char-acteristic of mountain areas of greatest elevationand relief. Table 2.3 provides examples of thesesubsystems and the main morphological unitswhich typify them.

Interactions between the coarse sediment sys-tem and fine sediment system in many mountainenvironments are difficult to separate as zonesof activity are closely coupled. For example, there

Geomorphology

Mountaingeomorphology

Mountainform

Morphodynamicsin mountains

Morphometryand structure

Relief generationand history

Morphoclimaticmodels

Process dynamicsand activity

Fig. 2.3 Research themes in mountaingeomorphology (Barsch & Caine 1984).Environmentalsedimentology is largelyfocused on themorphodynamics ofmountain environments but has important links to mountain form.(Reproduced withpermission MountainResearch andDevelopment.)

ESC02 9/8/06 11:24 AM Page 37

38 JEFF WARBURTON

is significant slope–channel coupling in mostmountain environments because, given the highrelative relief and steep slopes, valley sedimentstorage is often small and sediment runout from

mountain slopes enters stream channels directly.The usefulness of Fig. 2.4 is that it provides anoverall framework for evaluating sediment fluxesunder natural and disturbed conditions. Greater

Interfluve

Free-face

Talus

Talus foot

Valley floor

Stream channel

Lake basin

Bedrock weathering Atmospheric dust and solutes

Lake sedimentation Outlet channel

Internal transfers

Outputs

Inputs

Fig. 2.4 The alpine sedimentcascade process system proposedby Caine (1974).

Table 2.3 Examples of mountain geomorphological process subsystems and typical geomorphological units as described by Barsch &Caine (1984).

Sediment system

Glacia

Coarse debris

Fine sediment

Fluvial and geochemical

Morphological units

Glacierized valleysand terrain; moraineSteep bedrock slopesand talus

Waste mantled slopes

Stream channels;valley floors; fans andlakes

Transfer processes

Glacial transport

Rock fall, avalanches;debris flows; rockslides; talus creepSolifluction; soil creep;slopewashFluvial transport;solute transport; lake sedimentation

Typical mountainenvironment

Icelandic glaciers Ggjkull and KvrjkullRanda rock slide,Valais, Switzerland

Colorado FrontRange, USAKärkevagge,northern Sweden

Case study

Spedding (2000)

Gtz & Zimmermann(1993)

Benedict (1970)

Rapp (1960)

ESC02 9/8/06 11:24 AM Page 38


understanding can be achieved if the landformassociations in mountainous terrain can be shownin a land-systems model. Fookes et al. (1985)have developed such a model based on east Nepal,but as Gerrard (1990) argues, this model is moregenerally applicable to most extratropical highmountains that have been extensively glaciated.Figure 2.5 shows this model and the interrela-tionship between five main zones: high-altitudeglacial and periglacial; free rock faces and debrisslopes; degraded midslopes and ancient valleyfloors; active lower slopes; and valley floors.The model is most useful in demonstrating thecoupling between the slope sediment systems andvalley sediment systems that actively occur inhigh mountain environments and which form the

basis of conceptual models of sediment delivery(e.g. Fig. 2.4; Caine 1974; Barsch & Caine 1984).Given these different geomorphological subsys-tems there is an enormous variety of processesoperating in mountain environments. However,in terms of environmental sedimentology themost significant are usually the destructive massmovements which generally occur in mountaintorrents or on steep unstable bedrock slopes.

In terms of the fluvial system as a whole headwater mountain catchments are viewed as sediment production zones feeding bedloadand suspended sediment downstream. A simpleschematic model describing the linkages betweenmountain catchments and variations in fluvialform along the river profile from the headwaters

5

1

4

3

22

34

4

5

Colluvium

Residual soil orweathered rock

Braidedchannel with bar deposits

Taluvium cover toweathered bedrock

Fan

Alluvium

Young riverterrace deposits

Rock slopeRapids

Debris slide

Incised gully

Rotationalslide

Floodplaindeposits

Principalvalley

Active lower slopes

Ancient high levelterrace orerosion surface

Boulder field

High altitude glacialand periglacial

Degradeddebris slide

Degraded side slopeof former river valley

Rock facewith rockfall

Talus / screeTaluvium

Colluvium

Tributaryvalley

Fig. 2.5 Land-system diagram of a high-mountain environment showing five major terrain zones: (1) high-altitude glacial andperiglacial; (2) free rock faces and debris slopes; (3) degraded middle slopes and ancient valley floors; (4) active lower slopes; and (5) valley floors. (Redrawn from Engineering Geology, Fookes, P.G., Sweeney, H., Manby, C.N.D. & Martin, R.P. (1985) Geological andgeotechnical engineering aspects of low cost roads in mountainous terrain, 21, 1–152, with permission from Elsevier.)

ESC02 9/8/06 11:24 AM Page 39

40 JEFF WARBURTON

to the coastal lowlands was developed by Nevinin 1965 (Mosley & Schumm 2001; Fig. 2.6).This shows the transition from the headwatercatchments, which are dominated by steepchannels such as cascades, step–pool systemsand coarse braided rivers, through to meander-ing lowland channels. Generally the sedimentload in the headwater channels is dominated by coarse bedload (50–90% of the total load),whereas downstream fine suspended load dom-inates (often exceeding 80–90% of the totalannual load) (Fig. 2.6). It is well known that in the absence of significant tributaries grain-size systematically decreases downstream. Bedmaterial size at any point in the drainage basin isa function of sediment supply and the combinedaction of sediment sorting and abrasion. This is often expressed as an exponential decline ingrain size with distance downstream:

D = Do e−aL

where D is the particle size, Do is the initial particle size, L is distance downstream and a is

a coefficient representing the combined actionof sorting and abrasion. Such simple patterns do not hold for all rivers especially where thereare variations in different sediment source rocktypes or where tributary inputs of sedimentbecome significant (Rice & Church 1998).

2.2.2 Sediment yields from different mountainenvironments

The geomorphological activity in a particularmountain environment can be estimated by meas-uring the sediment yield from the catchmentsdraining such areas. There have been severalattempts to interpret global patterns of fluvialsediment yield (net erosion) in terms of the factorscontrolling sediment delivery (Walling & Webb1983; Jansson 1988; Milliman & Syvitski 1992;Summerfield & Hulton 1994). The main factorsconsidered have been either climate and runoff,or relief and tectonics. Clearly both are closelyrelated (see Fig. 2.2) and the sediment dischargefrom a particular drainage basin is dependenton the combination of these controls. A study by

Mountain/torrent phase Shingle phase Silt phase Tidal phase

Production zone Transfer zone Deposition zone

Hillslopesystem Fluvial system

Hillslo

pe

Colluvi

al

Cascade

Step–pool

Plane c

obble-g

rave

l bed

Braid

ed

Semi-b

raid

ed

Riffle

–pool

Freely

meanderin

g

Tidal

Fig. 2.6 Schematic diagram showing transitions in the fluvial system along a river profile. (Source: Mosley & Schumm 2001;reproduced with permission New Zealand Hydrological Society.)

ESC02 9/8/06 11:24 AM Page 40


Jansson (1988) is typical of this approach in that the research poses the question, ‘How doesthe magnitude of erosion vary globally underpresent human influence?’ The methodologicalapproach used in this study is a statistical analysisof sediment yield based on climatic conditions.No attempt is made to classify rivers in terms of topography. Alternatively, some studies haveexplicitly examined relief as a controlling vari-able on sediment discharge (e.g. Milliman &Syvitski 1992), whereas others have used a multi-variate approach incorporating morphometric,hydrological and climatic variables (Summerfield& Hulton 1994). The importance of drainage-basin topography in influencing mechanicaldenudation rates is clearly demonstrated in thestudy of Summerfield & Hulton (1994). Results(Table 2.4) show a relatively strong statisticalassociation between basin relief and mechan-ical denudation, albeit partly a function of other factors related to relief such as seismicity andweak rock structure.

Dedkov & Moszherin (1992) using suspendedsediment yield data from 1872 mountain riversassessed variations in erosion intensity andattempted to determine the significance of humanimpacts on mountain sediment systems. Theirdata (Fig. 2.7) show an interesting pattern thatsuggests all mountain areas are affected to somedegree by human activity but it is the areas oflower relief that are most greatly impacted. Fac-tors that promote erosion, such as forest removal,overgrazing, cultivation of slopes and road con-struction, occur in virtually all mountain regions

but the intensity of these activities varies withthe degree of economic development and popu-lation pressure. However, it is in the lower reliefmountains where these pressures are greatest.

Milliman & Syvitski (1992) in their analysisof 280 rivers discharging to the ocean found thatsediment loads/yields were a log-linear functionof basin area and maximum elevation of theriver basin. Other factors controlling sedimentdischarge, such as climate and runoff, were ofsecondary importance. In particular, sedimentfluxes from small mountainous rivers have beengreatly underestimated in previous global sedi-ment budgets, possibly by as much as a factor ofthree. Figure 2.8 shows the subdivision of river

Variable Log. mechanical denudation rate

Morphometric Area −0.11Mean trunk channel gradient 0.67Basin relief 0.80Relief ratio 0.78Mean model elevation 0.66Mean local relief 0.68Hypsometric integral −0.03

Hydrological Mean annual runoff 0.45Runoff variability −0.04

Climatic Mean annual temperature 0.41Mean annual precipitation 0.42

Table 2.4 Pearson correlationcoefficients for mechanical denudationversus morphometric, hydrological andclimatic variables. (Source: Summerfield& Hulton 1994.)

Landscapes close to natural

Modern landscapes

500

0

400

300

200

100

Lowlands Mountainshigh

Mountainsmiddle

Mountainslow

Rivers -source in the

mountains

Hills

Mea

n su

spen

ded

sedi

emen

t yie

ld (

t km

−2 y

r−1)

Fig. 2.7 Diagram showing the dependence of suspendedsediment yield on relief and the significance of humaninterference in enhancing erosion (Source: Dedkov & Moszherin1992; reproduced with permission of IAHS Press, from Dedkov,A.P. & Moszherin, V.I. (1992) Erosion and sediment yield inmountain regions of the world. In Walling, D.E., Davies, T.R. & Hasholt, B. (Eds.) Erosion, Debris Flows and Environment inMountain Regions. IAHS Publication 209, 29–36.)

ESC02 9/8/06 11:24 AM Page 41

42 JEFF WARBURTON

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

A: High mountain

Lad

x10

6(

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

E: Upland

Lad

x10

6(

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

B: Mountain Asia - Oceania

Load

106

)

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

F: Lowland

Load

106

)

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

C: Mountain N - S America, Africa, Alpine Europe

Load

x10

r)

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

G: Coastal plain

Load

x10

r)

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

D: Mountain non-Alpine Europe, high Arctic

Load

x1

yr)

Area x 106 (km2)

10000

0.001

1000

100

10

1.0

0.1

0.01

0.00001 0.001 0.1 10

Load

x1

6

Area x 106 (km2)

B C

D

E F G

Fig. 2.8 Relationship between sediment load and catchment area as classified by river basin category. Subdivision of river basins intofive categories based on maximum elevation within the hinterland: high mountain (headwaters at elevations > 3000 m), mountain(1000–3000 m), upland (500–1000 m), lowland (100–500 m) and coastal plain (< 100 m). The + symbols denote data points notincluded in regression calculations. (Source: Milliman & Syvitski 1992; reproduced with permission from The Journal of Geology,Milliman, J.D. & Syvitski, P.M. (1992) Geomorphic/tectonic control of sediment discharge to the ocean: the importance of smallmountainous terrain, 100, 525–44.)

ESC02 9/8/06 11:24 AM Page 42


basins into five categories based on maximumelevation within the hinterland: high mountain(headwaters at elevations > 3000 m), mountain(1000–3000 m), upland (500–1000 m), low-land (100–500 m) and coastal plain (< 100 m).Mountain rivers were further subdivided into:Asia and Oceania (generally very high sedimentyield), the high Arctic and non-Alpine Europe(low sediment yields) and the rest of the World.The correlations between load and basin areafor the various topographic categories generallyrange between 0.70 and 0.82. There is a distinctpattern to the data dependent on topographicsetting. Mountain rivers have the greatest loadsfollowed by the uplands, lowlands and coastalplains. There is some overlap in these generalrelationships owing to exceptions within each ofthe categories. For example, mountainous riversdraining South Asia and Oceania have muchgreater yields than (two to three times) othermountainous areas and are generally an order of magnitude greater than high Arctic and non-Alpine European mountains. Although thesestudies show some clear general patterns, suchdata should be interpreted with caution owingto the inherent errors in data collection and theincommensurate nature of the measurements(Harbor & Warburton 1993).

2.2.3 Sediment budget models of mountainsediment systems

Sediment budgets have been used as a tool for understanding sedimentary processes and

sediment fluxes for 50 years (Jäckli 1957, Table 2.5). A sediment budget accounts for the sources, transfers and storage of sedimentwithin a landscape unit. Constructing a contem-porary sediment budget for a mountain catch-ment is a time-consuming and labour intensiveendeavour and, therefore, most budgets tend to be measured over short periods (typically 1–3 yr). Figure 2.9 shows a sediment budgetframework applied to a small glacier basin insouthern Switzerland in order to evaluate thesignificance of the proglacial zone in contribut-ing sediment to a glacier-fed stream (Warburton1990). A measurement framework was set up to determine the rates of sediment transport of slope and channel processes and changes instorage within the sediment system (Fig. 2.9a).Results (Fig. 2.9b) clearly demonstrate theimportance of the glacier stream in terms of sediment flux but there is still a significant addi-tional load: approximately 23% of the totalcatchment sediment yield is added by proglacialsources (Warburton 1990). The overwhelmingproportion of the proglacial sediment (95%)was eroded from the valley floor during a briefmeltwater flood.

This methodological approach is widely appliedin the study of mountain sediment systems(Schlyter et al. 1993; Slaymaker et al. 2003). Twofurther examples of mountain sediment budgetmodels are shown in Table 2.5. These are fromthe Upper Rhine (area 4307 km2, relief 2800 m;Jäckli 1957) and Kärkevagge in northern Sweden(area 15 km2, relief 930 m; Rapp 1960). In terms

Table 2.5 Sediment fluxes from two high mountain environments: Upper Rhine (area 4307 km2, relief 2800 m; Jäckli 1957) andKärkevagge in northern Sweden (area 15 km2, relief 930 m; Rapp 1960). Units are in 106 J km−2 yr−1. A joule ( J) is the unit of work (E),which is generally defined by E = m g (h1 − h2), where m is mass, g is acceleration due to gravity and h is elevation, with (h1 − h2) beingthe change in height between two points 1 and 2. (Source: Barsch & Caine 1984.)

Catchment details Upper Rhine Kärkevagge

Area (km2) 4307 15Relief (m) 2800 930Coarse debris bedrock slopes 729.2 (4.2%) 15.7 (58.4%)Soil–fine sediment mantled slopes 53.5 (0.3%) 7.93 (29.5%)Channel transport–lake sedimentation 13,798 (79.5%) Not measuredSolute flux (output) 2781 (16.0%) 3.24 (12.1%)

Total 17,362 26.87

ESC02 9/8/06 11:24 AM Page 43

44 JEFF WARBURTON

Store

InterstreamInterstreamInterstreamwashwashwash

Gerlachtroughs

BluffCollapseCollapseCollapse

Channelstorage

Glacialsedimentsources

Moraine survey

Tributaries

Failure

Scour

Toe

Mainstream

Scour

Suspended sediment samplesBedload tracers

Tributary cross-profilesTributary long-profiles

ContinuousTurbidity

Electrical conductivity

DiscreteChannel survey

Bedload SamplesTracers

Suspended sediment samplesWater samples

Planform surveyLong-profilesPlanform photographs

Bank erosionCross-profiles

Sediment samplesBoulder survey

Gravel trapsSand trapsTurbidity

Electrical conductivity

Output

Bedload traps(basket / fence)

Basin output

Tributaries

Wash

Valley trainoutput

BluffCollapseCollapseCollapse

Glacier

HillslopeHillslopeHillslope

Storage

0.08

220.2

4792

21695

612561256125

560560560

131013101310

Moraine222.38

16399

11.411.411.4

108.1108.1108.1

Toestore

Duration of measurements25 May to 30 July 1987All values in tonnes

(a)

(b)

Hillslopes

store

Fig. 2.9 Sediment budget for Bas Glacier d’Arolla proglacial zone, Valais, Switzerland. (a) Diagram showing the sediment transferprocesses and storages, and measured sediment budget components. (b) Summary of main sediment fluxes and sediment storages inthe proglacial fluvial sediment budget May–July 1987 (values in tonnes). (Source: Warburton 1990.)

ESC02 9/8/06 11:24 AM Page 44


of sediment fluxes, regardless of the contrasts inscale, the coarse debris system (rock fall, snowavalanches and debris flows) is significant, andfluvial and channel processes are particularlyimportant (although this was not measured atKärkevagge). Slow mass movements tend to beof far less significance. Drainage-basin scale isimportant in determining the relative contributionof such processes because as basin size increasesvalley floor and channel processes become moresignificant (Church & Ryder 1972; Church &Slaymaker 1989).

The sediment transfer processes operating inthe Kärkevagge catchment are similar to thoseoperating in many mountain areas (Barsch &Caine 1984) (Table 2.6). The sedimentary char-acteristics of these processes reflect the sourcesediments within the catchment and the processdynamics that generate the subsequent sedi-mentation patterns. The overall picture is one of poorly sorted source sediments distributedacross a heterogeneous landscape. Differenti-ation of these deposits predominantly occurs asa result of gravitational sorting by rock fall,frost sorting of susceptible soil and selectivetransport by fluvial processes (Table 2.6).

When evaluating models of this kind it shouldbe kept in mind that Tables 2.5 and 2.6 and Fig. 2.9b show average values only for sedimentfluxes and neglect the inherent interannual vari-ability in such sediment systems and the longerterm dynamics of change that have an impact onmountain environments. In order to evalaute suchchanges longer-term sediment budget models needto be developed. Such models can be developedover historical time-scales using archival evid-ence (Piégay et al. 2004); the late Holocene usingdetailed lake sediment records and geochemicalanalysis (Slaymaker et al. 2003) and soil geomor-phology (pedology and weathering) (Birkelandet al. 2003). Furthermore, previous sediment-budget studies, such as the two examples providedin Table 2.5, can provide important baselinestudies for later comparison with further sedi-ment budgets or predictions of future change.Such an approach has been developed for Rapp’s(1960) work in Kärkevagge in northern Swedenby Schlyter et al. (1993).

Although the sediment budget framework isimportant for establishing how mountain sedi-ment systems operate and the relative import-ance of different sedimentological processes,

Table 2.6 Summary of slope denudation estimates and dominant sedimentary process characteristics in the Kärkevagge catchment1952–1960. (Based on: Rapp 1960.)

Process

Transportation of salts in running waterEarth slides and mudflows

Dirty avalanches

Rock falls

Solifluction

Talus creep

*The tons-metres vertical concept was introduced by Jäckli (1957) and is calculated by multiplying the mass of sediment moved by the verticalcomponent over which the sediment is transferred.

Sedimentary characteristics

Stream and lake solute loads derived from catchment-widechemical weatheringSources include poorly sorted till deposits and talus material.Deposition in bouldery mudflow levees, alluvial fans and sheetdeposits (sorted gravel and sand)Slush avalanches. Transported material ranging in size from finesilt to boulders up to 5 m in lengthIncludes: pebble falls, small boulder falls and large boulder falls(varying in size from 10 to 100 m3). Rock-fall debris dominantly20–50 cm, largest boulders up to 5 m in lengthStony soil with occasional boulders. Vegetated or partlyvegetated surfaces. Develops solifluction lobe movements 4 cm yr−1 to depths of 50 cmCoarse, clast-supported openwork surface with fine contentincreasing with depth. Poorly sorted surface sediments varyingin size from small pebbles to boulders. Surficial rates up to 10 cm yr−1

Ton-metres (vertical)*

136,500

96,375

21,850

19,565

5300

2700

ESC02 9/8/06 11:24 AM Page 45

46 JEFF WARBURTON

its usefulness for understanding environmentalsedimentology is limited because most of theimportant events are large-scale mass move-ments which are highly episodic and inevitablysite specific. Examples include localized flashfloods, rock slides and volcanically triggereddebris flows.

2.2.4 Sediment transfer processes

The magnitude and frequency of different sedi-ment transfer processes operating in mountainenvironments vary according to the local orregional tectonic and geological setting. It isbeyond the scope of this chapter to provide adetailed review of the environmental sediment-ology of volcanic, seismic and tectonic studies.However, the coincidence between these variablesand mountain areas is undeniable. For example,case studies later in this chapter clearly demon-strate links between seismic activity and debrisflows (Case Study 2.1 Huascaran, Peru); volcanic

activity and sedimentation (Case Study 2.2 MountSt Helens); and sediment delivery in tectonicallyunstable regions (Case Study 2.3 Waipaoa, EastCoast, New Zealand). Furthermore, these typesof events have common characteristics in termsof very high rates of sediment production oftencoupled to catastrophic or rapid releases ofwater or runoff. This leads to a consideration ofsediment–water flows because once a large massof water, ice, snow or sediment is released on aslope it will immediately begin to flow (Pierson1988). Such flows can rapidly entrain furthermaterial along their paths.

The behaviour of the flowing mass dependson the ratio of sediment to water. A very broadspectrum of flows is possible ranging from rela-tively dry (non-liquefied) granular flows of rockdebris (debris avalanches) to large flood flowsinvolving mostly water (Fig. 2.10). Eisbacher & Clague (1984) have classified destructivemass movements of this kind into five groups(Table 2.7):

G r a n u l a r m a t e r i a l

C o h e s i v e m a t e r i a l

Rapid motionSu

spen

sion

Bed load

Rapid motion SSSlllooowww mmmooottt iiiooonnn

Failure

Failure

Slow motion

TTTwww

ooo --- ppp hhh aaa sss eeefff lll ooo www

OOOnnn eee --- ppp hhh aaa sss eee fff lll ooo www

MudflowsMudflowsMudflows LandslidesLandslidesLandslides

Debris avalanches

StabilityStream flowsHyperconcentrated

flowsDebrisflows

(Granular flows)(Granular flows)(Granular flows)

Increasing water contentIncreasing water contentIncreasing water content

Increasing solid fractionIncreasing solid fractionIncreasing solid fraction

Water Solid

Rock falls

Fig. 2.10 Classification of mass movements and flows on steep slopes as a function of solid debris fraction and material type. (Source: Coussot & Meunier 1996; redrawn from Earth Science Reviews, Cousset, P. & Meunier, M. (1996) Recognition, classificationand mechanical description of debris flows, 40, 209–27, with permission from Elsevier.)

ESC02 9/8/06 11:24 AM Page 46


• debris flows from superficial deposits• debris flows from bedrock failure• mass movements on volcanoes• glacier-related mass movements• rock falls and rock avalanches

Recognizing the dominant flow type that occursduring a mountain flood event is an importanttask. Differentiation of the types of water andsediment flows, whether they are water floods,hyperconcentrated flows or debris flows, can be based on the degree of sorting, composition,texture and sedimentary structures (Table 2.8,Costa 1988). Water floods involve turbulentNewtonian fluid flow with non-uniform sedi-ment concentration profiles with sediment con-centrations less than 40% by weight. Debris flowson the other hand are rheologically very different.They are non-Newtonian having laminar flow anduniform sediment concentration gradients, withsediment concentrations varying from 70 to 90%by weight. Hyperconcentrated flows are transi-tional between these two extremes and, as such,retain some characteristics of both water floodsand debris flows (Costa 1988; Table 2.7).

Furthermore debris flows are often consideredas an intermediate phenomenon between hyper-

concentrated flows and landslides, both in termsof their initiation mechanism and dynamics.Coussot & Meunier (1996) propose a simplescheme for classifying mass movements andflows that occur on natural steep slopes. This isbased on material type and proportion of solid inthe moving mass (Fig. 2.10). Material types varyfrom fine, cohesive clays to coarse, cohesionlessgranular materials, and solid content generallyincreases from water flow, to hyperconcentratedflows, to debris flows and landslides. This is auseful summary of many of the main slope andvalley processes that operate in mountain envir-onments and how they are interrelated in termsof sedimentary continua, e.g. solid:water ratio(Table 2.7).

2.3 PROCESSES AND IMPACTS – EVENTS RELATED TO

NATURAL DISTURBANCE

This section will focus on natural changes inmountain sediment systems, particularly glacierand slope systems, that make up the coarse debrisand fine sediment components of the sedimentcascade model (Caine 1974).

Table 2.7 Examples of different types of sediment water flows and destructive mass movements in mountain environments.

Mass movement type

Debris flows from superficial deposits

Debris flows from bedrock failure

Mass movements on volcanoes

Glacier-related mass movements

Rock falls and rock avalanches

Flood

Location, date

GamaharaTorrent, Japan,1996

MonumentCreek, GrandCanyon, USA,1984

Mount St Helens,USA, 1980

Kautz Creek,Mount Rainier,Washington,USA, 1947

Elm, Switzerland,1881

Skei¨arársandurjökulhlaup,Iceland,1996

Origin

Landslide–debrisflow

Debris flow

Landslide

Debris flow(glacier outburstflood)

Rock avalanche

Glacier outburst

Volume (m3)

5 to 10 × 104

2.5 to 2.8 × 109

38 × 106

10 × 106

3.5 × 109

Velocity (m s−1)

16

70 (maximum)

4.5

50

5–10

Traveldistance (km)

4.6

4.5

25

9+

2

35+

Source

Marui et al.(1997)

Webb et al.(1988)

Pierson(1988)

Driedger(1988)

Heim(1882)

Magilliganet al. 2002

ESC02 9/8/06 11:24 AM Page 47

48 JEFF WARBURTON

2.3.1 Glacier systems and environmental change

A characteristic feature of many mountainousenvironments is the presence of glacier ice. Atpresent the distribution is restricted to the highermountains and polar ice sheets but in the pastwas considerably more extensive. The legacy ofpast glaciations still conditions sediment transferin most mountainous regions. This is manifest asa direct influence in currently glacierized regionsand as an indirect control in glaciated areas wherethe impact of glaciers still significantly altersmountain sediment systems. The concept of para-glaciation (Church & Ryder 1972) provides auseful framework for understanding contempor-ary mountain sediment budgets and the dis-equilibrium that often exists between sedimentproduction and delivery in previously glaciatedareas (Church & Slaymaker 1989). Ballantyne(2002) provides a comprehensive review of peri-glacial geomorphology and the important conceptof glacially conditioned sediment availability.He identifies six paraglacial land-systems andrecognizes the significance of these sedimentstores and sinks as sources of easily eroded sedi-ment. Sediment stores in mountain environments

are fundamental in controlling basin sedimentyield and may be sensitive to environmentalchanges in climate and/or human disturbance.In addition to these long-term influences glacialprocesses can have short-term effects on moun-tain sediment systems and in some cases posesignificant hazards. Three main hazards can beidentified: glacier fluctuations, glacier outburstfloods (jökulhlaups) and avalanches (Cooke & Doornkamp 1990). Hazards related to iceand snow are common in most glacierized highmountain areas. Their impact on society, how-ever, depends on the degree to which humanstructures and settlements are developed in thoseregions. The European Alpine countries are par-ticularly affected by glacier hazards owing tothe combination of steep, unstable slopes and theproximity of infrastructure and villages to theglacial environment. Of particular importanceis the sensitivity of the glacier environment tosmall changes in temperature and precipitation,which may considerably increase the risk forcommunities living near them. Outburst floods(jökulhlaups), landslides, debris flows and debrisavalanches can destroy property and take lives.Given the uncertainties of recent environmental

Table 2.8 Geomorphological and sedimentologic characteristics of water and sediment flows in channels (Source: Costa 1988;reproduced from Flood Geomorphology. Baker, V.R., Kochel, R.C. & Patton, P.C. (Eds). 1988. © John Wiley, New York. This material isused by permission of John Wiley & Sons, Inc.)

Flow

Water flood

Hyperconcentrated flow

Debris flow

*Trask sorting coefficient: calculated by dividing the 75th percentile by the 25th percentile of the grain size distribution. Φ graphic sorting:inclusive graphic standard deviation (in φ units) = [(φ84 − φ16)/4] + [(φ95 − φ5)/6.6].

Landforms and deposits

Bars, fans, sheets,splays; channels havelarge width:depthratios

Similar to water flood

Marginal levees,terminal lobes,trapezoidal to U-shaped channel

Sedimentary structures

Horizontal or inclinedstratification to massive;weak to strong imbrication;cut and fill structures;ungraded or graded

Weak horizontalstratification to massive;weak imbrication; thin gravel lenses; normal andreverse grading

No stratification; weak to no imbrication; inversegrading at base; normalgrading near top

Sediment characteristics*

Average Trask sorting coefficient1.8–2.7; clast-supported; normallydistributed; rounded clasts; widerange of particle sizes

Φ graphic sorting 1.1–1.6 (poor);clast-supported openwork structure;predominantly coarse sand

Average Trask sorting coefficient3.6–12.3; Φ graphic sorting 3.0–5.0(very poor to extremely poor);matrix-supported; negatively skewed;extreme range of particle sizes; maycontain megaclasts

ESC02 9/8/06 11:24 AM Page 48


change it is not clear yet whether some of thesehazards are a normal part of glacier behaviour,or whether they represent an evolving new threatfrom a changing cryosphere.

2.3.1.1 Fluvioglacial sedimentation and glacier outburst floods

A jökulhlaup is an Icelandic term used todescribe a catastrophic flood caused by the sudden drainage of a subglacial or ice-dammedlake. Lakes may develop and drain seasonally or may build up over many years before beingdrained. Volcanic eruptions or geothermal heat-ing under ice caps can often be the trigger forsuch periodic drainage, leading to catastrophicfloods.The Grímsvötn area on Vatnajökull insouthern Iceland is particularly prone to suchactivity, which is triggered by the Katla volcanounder the ice cap of Myrdalsjökull (Gerrard1990). At Grímsvötn water is stored in a sub-glacial lake located in a volcanic caldera of about30–40 km3 in the centre of Vatnajökull ice cap.Geothermal heating causes internal drainage inthe ice cap, which causes the subglacial lake to rise by about 100 m over a period of five orsix years. Eventually the ice dam is breached andwater flows under the ice cap to emerge at theice margin several tens of kilometres from thesource. Floods can have peak discharges of up to 100,000 m3 s−1. Two main mechanisms leadto the release of stored water. Drainage may beginat basal water pressures less than the ice over-burden pressure through the slow expansion ofglacial conduits by melting of ice walls throughfrictional or sensible heat. Alternatively highsubglacial lake levels lift the glacier off the bedalong the flow path resulting in extremely suddendischarges. As a consequence of the water releaseriver levels may rise by up to 10 m and millionsof cubic metres of sediment are deposited, oftenraising sandur (alluvial plain) levels by severalmetres. Sediments are transported from sub-glacial sources and eroded from the proximalzone of the moraine/sandur complex (Magilliganet al. 2002). Björnsson (2003) estimates the sediment load of a large jökulhlaup may be asgreat at 10 × 106 t per event, but if this occurs at

the same time as an eruption this may rise byanother order of magnitude. Similarly, Maizels(1997) distinguishes three types of proglacialsedimentary (outwash) deposits in relation tojökulhlaup type:

Type I – ‘normal’ braided river outwashType II – produced by sudden drainage of icedammed lakesType III – associated with drainage during subglacial geothermal activity, with depositsresulting from high sediment concentrationsand hyperconcentrated flows.

Each type of activity results in a distinct set of depositional landforms and sediments. In historical times, the Skeidarsandur jökulhlaupof November 1996 stands out as unprecedentedin its magnitude and duration (Magilligan et al.2002). The event reached a peak discharge of53,000 m3 s−1 in 17 hours and was responsiblefor widespread incision and aggradation with3.8 × 107 m3 of sediment being deposited in theproglacial depression of the sandur. Deposits of over 9 m depth were recorded. The total volume of water released from the glacier wasestimated at 3.5 km3, and was the most rapidjökulhlaup recorded for this area. The eventalso eroded large stretches of the main high-way around Iceland, destroyed two bridges, andcaused damage estimated at US$15,000,000(Magilligan et al. 2002).

Because glacier outburst floods are suddendischarges from water bodies dammed within or at the margins of glaciers in steep glacierizedenvironments the downstream impacts of suchevents can be devastating, leaving paths of totaldestruction. At Mount Rainier (4364 m) in theCascade Range, Washington State, USA glacialoutburst floods occur on a relatively small scale.Mount Rainier is a strato-volcano consisting of overlapping layers of lava and tephra. Themountain is topped by a summit ice cap, with 25 glaciers extending radially in all directionsfrom the summit (Fig. 2.11a). Most of the activ-ity on Mount Rainier is restricted to the glacierson south-western slopes of the mountain due tothe local geographical–climatological conditions

ESC02 9/8/06 11:24 AM Page 49

50 JEFF WARBURTON

46° 50' N

46° 55'

121° 50' W 121° 45' 121° 40'

MountMountMountRainierRainierRainier

Glacier IslandGlacier IslandGlacier Island

LongmireLongmireLongmireParadiseParadiseParadise

NNNiii sssqqquuuaaalll lll yyy RRRiii vvveeerrr

TTTaaahhhooommmaaa GGGlll aaaccciii eeerrr

SSS... TTTaaahhhooommmaaaGGGlll aaa

ccciii eeerrrPPPuuuyyyaaalll lll uuuppp GGGlll aaaccciii eeerrr

SSS... MMM ooowwwiii ccchhh GGGlll aaaccciii eeerrr

NNN... MMMooowwwiii ccchhh GGGlll aaaccciii eeerrr

KKKaaauuu

ttt zzz

GGGlll aaaccciii

eeerrr NNNiii sssqqquuuaaalll lll yyy

GGGlllaaaccciiieeerrr

CCCooowwwlll iii ttt zzzGGG

lll aaaccciii eeerrr

III nnngggrrr aaammm

GGGlll aaaccciii eeerrr

EEEmmmmmmooonnnsss GGGlll aaaccciii eeerrr

WWWiii nnn

ttt hhhrrr ooo

pppGGG

lll aaaccciii

eeerrr

CCCaaarrr bbbooonnnGGG

lllaaaccciiieeerrr

KKK aaa uuu tttzzz

CCCrrr

eeeeeekkkCCCrrr eeeeeekkk

TTTaaahhhooommmaa

0 5kilometres

Mt. RainierMt. RainierMt. Rainier

(a)

(b)

Fig. 2.11 (a) Location map of Mount Rainier, WashingtonSate, USA showing principal glaciers. (b) Tree damage caused bydebris flow activity in Kautz Creek (1989).

ESC02 9/8/06 11:24 AM Page 50


favouring subglacial cavity formation: increasedexposure to storms and more intense solar radiation (melt) (Driedger & Fountain 1989). AtMount Rainier water originates from snow andice melt and liquid precipitation, and is storedwithin glacier cavities and at the bed of the glacier.Ice movements deform these cavities, resultingin the catastrophic release of the water. Hydro-logical studies of the South Tahoma glacier(Driedger & Fountain 1989) suggest that wateris stored at the bed of the glacier and the floodmagnitudes originating from such storage are inthe order of 1 × 105 m3. These floods threaten lifeand property because they occur without warningand quickly develop into rapidly moving debrisflows as they entrain loose volcanic debris.

Kautz Creek (Fig. 2.11a) is the catchmentaffected by the largest recent event, which occurredon 2–3 October 1947. This was triggered byheavy rain and it was estimated that 38 × 106 m3

of material was moved in the event, with someboulders up to 4 m in diameter being trans-ported (Driedger 1988). The highway 9 kmdownstream from the glacier was engulfed bythe flow, which deposited over 3 m of mud anddebris. Since 1947 smaller significant floods haveoccurred in 1961, 1985 and 1986. Tahoma Creekshows clear evidence of frequent outburst floodsfrom South Tahoma glacier. At least 22 outburstfloods have been recorded since 1967, including14 in the years 1986–1992 (Walder & Dreidger1995) (Table 2.9). In 1967, 1971, 1973, 1986and 1988 small floods triggered debris flows thatdestroyed trail bridges and campground-picnicareas. Recent activity has resulted in valley scourof 2 m in places, with boulders up to 0.5 mtransported and local deposition exceeding 1 m

in places. On the Nisqually River (Fig. 2.11a),outburst flooding from the Nisqually glacierdestroyed and damaged bridges in 1926, 1932,1934 and 1955, resulting in a high-level bridgeeventually being built. Since then large floods in1968, 1970, 1972 and 1985 resulted in massiverearrangement of the stream bed but have leftthe bridge unscathed. Other activity has also beennoted at the Winthrop glacier, and the Carbon,South Mowich and Emmons glaciers are alsosuspected of being susceptible to this kind ofactivity (Fig. 2.11a). The most recent activityreported from Mount Rainier was on 14 August2001 and constituted a moderate debris flow in the Van Trump drainage within the KautzGlacier area.

The outburst floods tend to occur in late sum-mer or autumn; usually in the late afternoon or early evening, often during rainstorms. Theflood waves have some common characteristics.They have been described as a noisy, churningmass of mud and rock. Local winds can developalong the flows and thick dust clouds can accom-pany the events. There is often a smell of freshlycut vegetation and chipped rock as boulderssmash trees and collide with bedrock and othersediment (Fig. 2.11b). The flows are very rapid,often in excess of 4.5 m s−1. Observers have saidthere is generally less than 2 minutes betweenhearing the flow and it passing the observer(Driedger 1988) (Table 2.7).

2.3.2 Seismically triggered slope instabilities

Landslides or rock falls occur when a surficialmass fails along a steeply dipping fracture plane.The rock mass breaks up and moves downslope,

Glacier Surface area Mean surface slope Number of floods in (km2) (degrees) record up to 1988

Nisqually 4.6 25 9South Tahoma 2.8 23 12*Kautz–Success 1.8 29 5Carbon 11.2 18 1Winthrop 9.1 21 2

*Revised estimate (Walder & Driedger 1995) 22 floods since 1967.

Table 2.9 Characteristics of glacierssusceptible to outburst floods – Mount Rainier, Washington, USA.(Source: Driedger & Fountain 1989.)

ESC02 9/8/06 11:24 AM Page 51

52 JEFF WARBURTON

breaking into fragments and then continues as aslide or flow. Such mass movements have severalcommon characteristics:1 landslides occur in unstable, faulted andjointed rock masses;2 a trigger mechanism is usually involved in setting them off;3 flow within the landslide is complex – ofteninvolving several stages or modes of movement;4 once deposited the mass is often still un-stable and undergoes modification by other slopeprocesses;5 these are very rapid mass movements.

Seismically triggered slope instabilities havemany of these features (see Case Study 2.1). Forexample, the Sherman Glacier rock avalanche in Alaska was triggered by the 1964 AlaskaGood Friday earthquake (McSaveney 1978). Thebedrock of this area consists of highly deformedslightly metamorphosed sedimentary sequences,which are heavily faulted and the fault planes aresteeply dipping. The rock avalanche fell 600 mand then spread 5 km across the Sherman Glacier,depositing a blanket 3–6 m thick. The avalancheflowed as a large lobate mass with the particlesbehaving like loose aggregates, spreading as it

Case study 2.1 Huascarán, Yungay, Peru 1970

The greatest mountain landslide disaster recorded to date occurred in Peru following an offshoreearthquake (magnitude 7.7) on 31 May 1970. The seismic event triggered a catastrophic rockand snow avalanche from the summit ice-cap of Huascarán Mountain (6654 m), the highest peakin the Peruvian Andes. The displaced mass fell vertically to the glacier below, where it generateda gigantic debris flow surge that travelled down-valley at a speed of over 70–100 m s−1 entrain-ing sediment in its path (Case Fig. 2.1). On reaching the meltwater river of the Rio Sacsha theflow had changed into a mudflow of over 1 km wide, carrying gigantic boulders, some as greatas 15 m3. Eyewitnesses described the flow as a huge wave at least 80 m high (Whittow 1980).The flow travelled 15 km down to the confluence with the Rio Santa in just a few minutes.

The mountain has had a history of devastating landslides. A mudflow had swept down the RioSacsha in 1962 killing 4000 people and depositing 13 × 106 m3 of material (Smith 2001). The1970 mudflow was of much greater magnitude and completely overran the towns of Yungay andRanrahirca, leaving 18,000 people dead or missing. It is estimated that 50–100 × 106 m3 ofmaterial was involved, burying some settlements up to 10 m deep in sediment (Case Table 2.1).

This is not the only catastrophic mass movement event identified in this valley. A smalleravalanche also took place on 10 January 1962 and an earlier (Pre-Columbian) event, larger thanthe 1970 avalanche, is identified from deposits preserved on the valley floor (Case Table 2.1).

Case Table 2.1 Summary characteristics of selected data from three of the largest avalanches observed from NevadosHuascarán, Peru. The Pre-Columbian event is identified from deposits preserved outside the limits of the 1970 event. (Source: Plafker & Ericksen 1978.)

Pre-Columbian 10 January 1962 31 May 1970

Area covered (km2) > 30 6 22.5Volume (106 m3) 100–200* > 13 50–100Average velocity (km h−1) 315–355† 170 280Runup height at Rio Santa (m) 123 30 83Velocity at Rio Santa‡ (km h−1) > 140 > 60 > 120

*Estimated from the extent and nature of deposits.†Derived by comparison with historic avalanches.‡Calculated from runup height assuming avalanche front thicknesses of 15 m (1962), 30 m (1970) and 45 m (Pre-Columbian).

ESC02 9/8/06 11:24 AM Page 52


Although details of the Pre-Columbian event are less well known and no estimates of loss of lifeare available, the 1962 event destroyed nine towns including most of Ranrahirca (Case Fig. 2.1),killing approximately 4000 people. Although the trigger mechanism for this event is not known theavalanche appears to have originated as an icefall from the same part of the west face of the NorthPeak of Huascarán. Based on evidence from deposits this was predominantly a rock avalanche.

II

I

456245624562

418341834183390739073907

538153815381

332433243324

313331333133

309530953095

280028002800294029402940

275527552755278927892789

253825382538

665466546654HuascaranHuascaranHuascaran

RanrahircaRanrahircaRanrahirca

Yu ngayYu ngayYungay

MalacotoMalacotoMalacoto 0 km 2

NRRR iii ooo SSS aaa nnn ttt aaa

RRRiii ooo

SSShhh

aaaccc

sss hhhaa

a

GlacierGlacierGlacier

I I I I I I

I I I I I I I I

Villages / TownsVillages / TownsVillages / Towns

RoadsRoadsRoads

1970 avalanche1970 avalanche1970 avalanche

1962 avalanche1962 avalanche1962 avalanche

Heights in metresHeights in metresHeights in metres538153815381

RiversRiversRivers

a

Case Fig. 2.1 Map of the Rio Shacsha and NevadoHuascará area, Peru showing therunout tracks of the 1962 and1970 avalanches. (Modified fromWhittow 1980.)

ESC02 9/8/06 11:24 AM Page 53

54 JEFF WARBURTON

went. Because of the high speed of this flow andthe mechanical fluidization through jostling ofthe clasts, flow was very plastic (high-viscosity–low-yield stress). High viscosity prevented turbu-lence and the flow was almost laminar. Becauselittle energy was used up in internal deformation

most energy was lost to friction at the base,thereby eroding the substrate.

Another example is the 1980 Mount St Helensdebris avalanche in Washington. This was causedby the eruption of 18 May (Case Study 2.2). Itbegan catastrophically with a large lateral blast

Observation of the summit area of Nevados Huascarán suggests that there is still consider-able potential for further destructive avalanches. Slopes in the source area remain oversteepenedand show signs of recent cracks parallel to the avalanche scar. Therefore the potential for furtherdisasters remains. Although the settlements are nearly 23 km from the source of the debris flows,given the steepness of the mountain slopes, travel times are less than four minutes (Plafker & Eriksen 1978). Even if a warning could be given evacuation to safe ground in such a shorttime period is not a viable option. Therefore a more viable option would be the relocation ofsettlements in the Santa valley to positions outside the maximum extent of the Pre-Columbianavalanche. This, however, would require a major initiative in development planning.

Relevant reading

Plafker, G. & Eriksen, G.E. (1978) Nevados Huascaran avalanches, Peru. In Rockslides and Avalanches,Vol. 1. Natural Disasters (Ed. B. Voight), pp. 48–55. Elsevier, Amsterdam.

Smith, K. (2001) Environmental Hazards – Assessing Risk and Reducing Disaster, 3rd edn. Routledge, London.Whittow, J. (1980) Landslides and avalanches – avalanches. In Disasters: the Anatomy of Environmental

Hazards, pp. 163–70. Penguin, Harmondsworth.

Case study 2.2 Impact of an extreme tectonic/volcanic event on a mountain sediment system –Mount St Helens eruption 1980

The 18 May 1980 eruption of Mount St Helens in south-west Washington was a major geo-logical disaster claiming 57 lives and affecting several hundred thousand people. A magnitude5+ earthquake triggered a lateral blast, which sent a huge mass failure down the northern flankof the mountain eventually exploding in a cloud of ash, rock and hot gas, sending ash up to 18 km vertically into the atmosphere. The avalanche of rock, ice and mud released in the earth-quake surged 28 km down the North Fork Toutle River valley (Case Fig. 2.2a & b). A secondpart of the slide flowed into South Coldwater Canyon and Spirit Lake raising the lake level by70 m and damming the outlet with debris over 100 m deep. Downstream mudflows choked theCowlitz and Toutle rivers bringing shipping to a halt on the Columbia River (Case Fig. 2.2a).

The immediate engineering response to the Mount St Helens disaster was undertaken by theUS Army Corps of Engineers, who were responsible for navigation and flood control on theColumbia River. The worst affected area was the Cowlitz River on highway I5 where effortswere concentrated on raising river levees and roads and clearing channel constrictions (CaseFig. 2.2a). Overnight the navigation channel on the Columbia River was reduced in depth from 12 to only 4.2 m. Using large dredges by 23 May the channel was partially cleared and by

ESC02 9/8/06 11:24 AM Page 54


Mount St HelensMount St HelensMount St Helens

WashingtonWashingtonWashington

10000

0.01

1000

10

0.1

1980 200019961984

Sed

imen

t yie

ld, i

nth

ousa

nds

of M

g km

−2

LongviewLongviewLongview

1998 1992

1.0

100

Water year

SRS

KelsoKelsoKelso

Castle RockCastle RockCastle Rock

CougarCougarCougar

Kid Va lleyKid Va lleyKid Valley

Si lverSi lverSilverLakeLakeLake Spi r i tSpi r i tSpirit

LakeLakeLakeCastleCastleCastle

LakeLakeLake

RRRiii fff fff eee LLL aaakkkeee

Merr i l lMer r i l lMerrillLakeLakeLake

YaleYaleYaleLakeLakeLake

Swi ftSwi ftSwiftReservoi rReservoi rReservoir

CCCooolll uuummm

bbbiii aaaRRR

iii vvveeerrr

CCCooowwweeeeeemmmaaannn RRRiii vvveeerrr

CCCooowwwlll iii ttt zzzRRRiii vvveeerrr

SSSooouuuttt hhh FFF ooorrr kkk TTTooouuuttt lll eee RRRiii vvveeerrr

NNNooorrr ttt hhh

FFFooorrr kkk TTTooouuuttt lll eee RRRiii vvveeerrr

GGGrrr eeeeeennn RRRiii vvveeerrr

TTTooouuuttt lll eeeRRRiii vvv

eeerrr

KKK aaalll aaammmaaa RRRiii vvveeerrr

PPPiii nnn

eeeCCC

rrr eeeeeekkk

MountMountMountSt HelensSt HelensSt Helens

46° 00' N 123° 00' W123° 00' W123° 00' W

46° 15’

46° 30’

123° 30'123° 30'123° 30' 123° 15'123° 15'123° 15' 122° 00'122° 00'122° 00'

CCCrrr eee

eeekkk

SSSwww

iii fffttt

LLLeee

wwwiii sss

RRRiii vvv

eeerrr

MMMuuudddddd

yyyRRR

iii vvveeerrr

0 10kilometres

(a)

(b) (c)

Gauging station Area altered by debris avalanchefrom the 18 May 1980 eruption

North Fork Toutle River (KID)Muddy River (MUD)South Fork Toutle River (SFT)

Green River (GRE)Toutle River (TOW)

Case Fig. 2.2 Impacts of the 1980 Mount St Helens eruption. (a) Location map showing main river systems. (b) North ForkToutle River Valley, 1989. (c) Annual suspended sediment yields in Mount St Helens rivers. The vertical line (SRS) is the locationof the sediment retention structure. The horizontal line is mean sediment yield value for western Cascade Range rivers (Source: Major et al. 2000; reproduced with permission Geological Society of America.)

ESC02 9/8/06 11:24 AM Page 55

56 JEFF WARBURTON

November the shipping lane was fully restored. Temporary dams and sediment storage pitswere excavated in the winter of 1980 in the Toutle River to hold back sediment delivery fromthe upper catchment, and dredging operations were completed in May 1981. These temporarystructures have now been overwhelmed by aggradation of the river (Case Fig. 2.2b).

By 1982 lake waters at Spirit Lake were rising dangerously high behind the eruption debrisdam. By 19 August 1982 a State of Emergency was declared at Spirit Lake and a full-scaleFederal response was initiated. In November 1982 water was being pumped from Spirit Lake toimmediately reduce water levels, however, in order to permanently drain the lake to safe levelsa water transfer tunnel connecting the lake to South Coldwater Creek was built and this wasoperational by 6 May 1985.

In order to produce a long-term solution and management strategy for the sediment prob-lem the Mount St Helens Sediment Retention Structure (SRS) was designed and built. In 1986work began on the construction of the SRS on the North Fork Toutle River. The purpose of thestructure was to stop the advance of debris avalanche deposits from the 1980 volcanic eruptionmoving downstream and causing long-term navigation problems on the Toutle, Columbia andCowlitz rivers. This was one element in a three part solution of long-term sediment management;this also involved river levee construction and a dredging programme. The US Army Corps ofEngineers initiated the project in 1986 and impoundment began in November 1987. The aim of the sediment retention structure was to dam sediment not water. The structure reduces theflow to such an extent that sediment is deposited naturally in the upstream side of the structureand does not migrate downstream causing flooding problems and shipping hazard. The SRSconsists of a 600 m wide embankment standing nearly 56 m above the pre-eruption streambed. The embankment is constructed of crushed and fractured rock with an impervious core ofclay. The embankment rests on gravel, and water passes underneath and into the embankmentas the lake rises. The outlet works occupy a large central block, which consists of six rows offive outlet pipes through which water and fish can pass into the plunge pools and downstreamoutlet channels. The pipes are closed off permanently as the level of sediment rises in theimpounded lake upstream. Once the conduits are fully blocked the river will flow continuouslyover the wide unlined spillway, which is approximately 38 m above the original stream bed.The structure is sited about 35 km downstream of the volcano.

The impounded area is estimated to have a sediment retention capacity of approximately 198 × 106 m3 of sand and gravel and is expected to fill in 50 years. In 2000 the structure wasapproximately 30% full. The rate of sediment retention varies greatly from year to year. Forexample, in the wet winters of 1996 and 1997, the higher than average discharges resulted in 23.5 × 106 m3 of sediment being deposited behind the structure. This was nearly four timesthe amount trapped in the previous two years.

Relevant reading

Major, J.J. (2003) Post-eruption hydrology and sediment transport in volcanic river systems. Water ResourcesImpact 5, 10–15.

Major, J.J., Pierson, T.C., Dinehart, R.L., et al. (2000) Sediment yield following severe volcanic disturbance – atwo decade perspective from Mount St Helens. Geology 28, 819–22.

Major, J.J., Scott, W.E., Driedger, C., et al. (2005) Mount St. Helens erupts again: activity from September 2004through March 2005. U.S. Geological Survey Fact Sheet FS2005-3036. http://vulcan.wr.usgs.gov/Volcanoes/MSH/Publications/FS2005-3036/FS2005-3036.pdf (accessed July 2005).

ESC02 9/8/06 11:24 AM Page 56


to the north of the mountain followed by theeruption. Erupted debris and the collapse of the north side of the mountain, together withmelting of the summit icefields, resulted in pyro-clastic flows off the summit cone (lateral flow of hot gases and unsorted volcanic fragments).This initial surge deflated and was translatedinto a ground-based debris flow, which continuedas a debris avalanche down the North Fork of the Toutle River. This became a debris flow andgradually evolved into a hyperconcentrated flowas it moved downstream (Tables 2.7 & 2.8).

2.3.3 Natural and climatically induced slope failures

Mass wasting of slopes in mountain environ-ments proceeds by a combination of small-scaleprocesses and infrequent large-scale events (seeCase Study 2.3). In the Swiss Alps debris flowsare a widespread phenomenon occurring in allaltitudinal belts, although they are especiallycommon in the periglacial zone. Debris flow start-ing zones tend to occur in poorly consolidateddebris on slopes or in gullies. The association

Case study 2.3 Impact of an extreme climate event on the hillslope sediment system – Cyclone Bola(New Zealand)

In 1998 Cyclone Bola had a major impact on much of the North Island, New Zealand with theworst affected area being on the east coast (Case Fig. 2.3c). Cyclone Bola was the largest stormevent in the Waipaoa catchment that had occurred since European settlement in the 1830s.Between 6 and 9 March up to 900 mm of rain fell in the north-east of the catchment (Page et al.1999). The Waipaoa is soft-rock (Cretaceous greywackes and Miocene–Pliocene siltstones andsandstones) hill country on the east coast of the North Island, New Zealand (Case Fig. 2.3b).This is a tectonically active area with an uplift of approximately 3 mm yr−1. The storm was estimated to be a 100 year return period event. The main impact of the storm was extensive erosion of hillslopes and massive sediment transfer along the valley systems. This is not the onlymajor erosional event in this area. Deforestation and conversion to pasture in the wake ofEuropean settlement initiated a major phase of instability in the catchment. Similarly largestorms over the past 100 years have also had major impacts on erosion and sedimentation inthe watershed (Trustrum et al. 1999).

A short-term sediment budget was constructed by Page et al. (1994) to assess the response ofCyclone Bola on the smaller Tutira catchment (3208 ha) to the intense rain storm event (CaseFig. 2.3a). The budget quantifies the total amount of sediment generated during the event andthe relative contribution from different erosion processes. Sediment storage is also estimatedalong with the amount of sediment discharge into two lakes within the catchment. A total of1.35 (± 0.13) million cubic metres of sediment was moved during the storm (420 m3 ha−1). Of thistotal, 21% was stored on the hillslopes, 22% deposited on the valley floors, 51% was depositedin the lakes and the remaining 6% was discharged at the catchment outlet. Approximately 89%of the sediment generated during the storm was from landslide erosion on the slopes. Channel,gully and sheet erosion was only a minor component of the budget (Case Fig. 2.3d).

Because of the dominance of landslides in the Cyclone Bola event work has been undertakento characterize the significance of this in sediment delivery (Page et al. 1999; Trustrum et al.1999). Page et al. (1999) developed a method for assessing sediment production from land-sliding in the Waipaoa catchment and applied this to the Cyclone Bola event with the aim of determining the contribution of landslides to suspended sediment output from the event(Case Fig. 2.3b). Using a geographical information system (GIS) containing the distribution of

ESC02 9/8/06 11:24 AM Page 57

58 JEFF WARBURTON

landslide terrain units, vegetation and storm characteristics, relationships established betweenstorm rainfall and landslide frequency were used to estimate landslide density across the entire2205 km2 catchment (Case Table 2.3). Each land system group has a characteristic rock type,suite of landforms and erosion processes, drainage density and channel morphology. Sixteenhave been identified but only the six listed in Case Table 2.3 are prone to landslides. The TeArai land system is most prone to landsliding, being made up of very weak bedrock, steepslopes with broken surface soil structure and steeply incising channels.

36°10'S

176°50'E176°50'E176°50'E 178°E178°E178°E

36°30'S

Poverty Bay

N

Hawkes BayHawkes BayHawkes Bay

Ta sma n SeaTa sma n SeaTasman Sea

Pacif ic OceanPacif ic OceanPacif ic OceanPacif ic OceanPacif ic Ocean

WWW

TTT

(a) (b)

(c)

LakeLakeLakeTuti raTuti raTutira

0 km 2 0 kilometres 20

Landslide-prone Steepland

Landslide-prone Steepland

0 km 200

(d)

Case Fig. 2.3 (a) Tutira catchment; (b) Waipaoa River basin; (c) location of the two study areas (T, Tutira; W, Waipaoa) in North Island, New Zealand; and (d) Lake Tutira and the adjacent landslide-prone slopes, showing the 1988 Cyclone Bola storm-triggered landslides and floodplain sedimentation (photograph courtesy of N. Trustrum). (Source: Trustrum et al.1999; reproduced with permission Gebr. Borntraeger Verlagsbuchh.)

ESC02 9/8/06 11:24 AM Page 58


between debris flow activity and ground ice meansthat the formation of debris flows is susceptibleto climate change, especially in marginal peri-glacial areas (Zimmermann & Haeberli 1992).In particular, since the Little Ice Age (ad 1450to 1890) glacier retreat, disappearance of per-ennial snow patches and degradation of low-lying permafrost have resulted in the exposure ofthawed debris on steep unstable slopes. Continuedwarming will result in an expansion of this zone.Changing climate also influences storm charac-teristics and changes the temporal and/or spatialpattern of precipitation and snow deposition.

An increased incidence of large-scale debris flowsfrom within the Alpine periglacial belt poses athreat to mountain villages and infrastructure asmany debris flows will be large enough to reachthe valley floor.

In the summer of 1987 the Alpine coun-tries were devastated by a number of floods. InSwitzerland eight people were killed and damageexceeded 1.3 billion Swiss Francs (c. 1 billionUS$). Heavy summer thunderstorms followed a long winter of extended snow cover and acold/wet spring. The worst affected areas weresouth-east Switzerland, on 18–19 July and again

Shallow landslides are responsible for approximately 64% of the load at the exit from thecatchment. In respect of longer term suspended sediment yield, however, landslides contributeonly 10–19% of the load (Page et al. 1999). Erosion of stored sediment in tributaries is animportant control on longer term sediment production (Marutani et al. 1999). In the few yearsimmediately following an event of this magnitude suspended sediment concentrations are 100%greater than in the years preceding the event owing to continued erosion of landslide scars andstored sediment (Marutani et al. 2001).

Relevant reading

Marutani, T., Brierley, G.J., Trustrum, N.A., et al. (2001) Source-to-sink Sedimentary Cascades in Pacific RimGeosystems. Matsumoto Sabo Works Office, Ministry of Land, Infrastructure and Transport, Japan.

Page, M.J., Trustrum, N.A. & Dymond, J.R. (1994) Sediment budget to assess the geomorphic effect of acyclonic storm, New Zealand. Geomorphology 9(3), 169–88.

Page, M.J., Reid, L.M. & Lynn, I.H. (1999) Sediment production from Cyclone Bola landslides, Waipaoa catch-ment. Journal of Hydrology (New Zealand) 38(2), 289–308.

Trustrum, N.A., Gomez, B., Page, M.J., et al. (1999) Sediment production, storage and output: the relative role of large magnitude events in steepland catchments. Zeitschrift für Geomorphologie N.F. Supplement115, 71–86.

Case Table 2.3 Summary of sediment production and delivery produced from landslides in the Waipaoa catchment (North Island, New Zealand) during Cyclone Bola 1988. Estimates based on storm isohyet data. (Source: Page et al. 1999.)

Land system Percentage area of Percentage Sediment generation Percentage of total Percentage contribution tototal catchment pasture* (m3)† suspended sediment load

Te Arai 23 96 20,053,000 61 39Wharerata 12.5 75 6,809,000 21 13Waihora 3.5 76 577,000 2 1Wharekopae 20.5 92 1,978,000 6 4Makomako 6 84 2,493,000 7 5Waingaromia 1.5 89 948,000 3 2Total 148,480 ha 32,858,000 100 64

*As of 1988 – excludes indigenous forest, scrub and exotic forest > 8 years old.†Adjusted for forest, scrub and soil conservation plantings.

ESC02 9/8/06 11:24 AM Page 59

60 JEFF WARBURTON

on 24–25 August. Over 600 debris flows occurredduring this period and because most of the pre-cipitation fell as rain even at high altitudes, over60% of the debris flows, including 82 large events(volumes > 1000 m3), originated above 2300–2400 m a.s.l. in the periglacial belt; on the lowerslopes conditions remained stable. Rainfall inten-sities were relatively low, however the period ofprecipitation was long and most debris flows werereported during or shortly after the maximumdownpour. Starting zones were typically 28° to33° for slope-type debris flows, but a substantialproportion (25%) were initiated in torrents,gorges and rocky ravines (Haeberli et al. 1990;Rickenmann 1990; Zimmermann & Haeberli1992; Rickenmann & Zimmermann 1993).

The event has been investigated in detail by an interdisciplinary team of Swiss scientistsexamining the causes and impacts of the storms(Haeberli et al. 1990).1 Debris flows were initiated in thick and extre-mely loose debris and bedrock. No sedimentarystructure could be considered a factor that limitedthe depth of erosion. Erosion depth was morelikely caused by the dynamics of the flow. Thesediments affected appeared to be highly per-meable and hydraulically non-homogeneous.The widespread occurrence of such material instarting zones makes these areas prone to fur-ther instabilities (Haeberli et al. 1990).2 The regions affected by the debris flows aremainly located in crystalline bedrock covered in massive and stratified, loose glacial and talusdeposits with poor sorting and low clay and limited silt contents (Roesli & Schindler 1990).3 Most of the debris flow activity in 1987occurred on unstable slopes or in gullies pre-disposed to instability. The spatial distributionof debris flows shows a distinct concentration in the periglacial zone as well as small clusters in small tributary valleys. The occurrence of the events could not be predicted from a knowledge of meteorological conditions alone(Zimmermann 1990).

Zimmermann & Haeberli (1992) did a comparison of the probability of permafrostoccurrence in the 1987 debris-flow startingzones and conditions at the same locality in

the Little Ice Age. Table 2.10 indicates that the 1987 debris flows predominantly occurredin localities where permafrost is marginal orabsent. Significantly in about 20% of cases con-ditions over the historical period changed fromprobable to possible permafrost occurrence andin such areas permafrost degradation and thawinstability are likely to have taken place prior tothe 1987 debris flows.

Another form of mass movement which canhave a large impact in developed mountain coun-tries are massive rock-slides. The 1991 Randarock slide in southern Valais, Switzerland causedextensive environmental damage and a heavyeconomic burden. The rock slide occurred in the tourist valley leading to Zermatt and theMatterhorn (Fig. 2.12). Approximately 30 ×106 m3 of rock fell from a south-east facing rockface near the village of Randa approximately 10 km north of Zermatt. The rock fall occurredin two main parts: on 18 April 1991 and againon 9 May 1991. There were no fatalities exceptfor loss of some livestock and 31 chalets wereburied in the events (Quanterra 2003). The firstevent occurred without warning and caused majordisruption and devastation to the valley infra-structure. The first event deposited approxim-ately 20 × 106 m3 of rock and initial assessmentsshowed there was still a considerable mass ofunstable rock in place. This prompted authoritiesto install a seismographic and geodetic warningsystem. As a result the second event was forecastfrom field evidence and geodetic and seismicsurveys, and the area was successfully evacuated.Prior to the second slide geodetic displacement

Table 2.10 Comparison of the probability of permafrostoccurrence in the 1987 debris flow starting zones forcontemporary and Little Ice Age conditions in the Swiss Alps.(Source: Zimmermann & Haeberli 1992.)

Occurrence of Percentage of all cases (n = 82)permafrost in areas of 1987 debris flows 1850 1987

Certain 6 6Probable 27 9Possible 23 21Unlikely 44 64

ESC02 9/8/06 11:24 AM Page 60


measurements indicated an accelerated motionof the rock mass (Götz & Zimmermann 1993)(Fig. 2.12b). During the evening of 9 May 10 ×106 m3 of rock was deposited in a series of slides(generally < 1 × 106 m3) over a period of 7 hours.Deposits covered approximately 0.8 km of theZermatt–Rhône Valley railway line and about0.2 km of the main road. The deposit caused a rock-slide dam across the Vispa River

resulting in 30 houses being flooded. Dust wasdeposited to a depth of 10–40 cm in a 1 kmradius around the slide area and some localhousing was buried in debris up to 60 m deep(Götz & Zimmermann 1993).

Formation of the rockslide-dammed lake inthe valley posed two major hazards: local flood-ing and a sudden dam break flood downstream.Remedial measures were taken to keep lake levels

2.0

0

1.5

1.0

0.5

15 151052520 30April May

Hei

ght (

met

res)

N

R hone

VB

MMMaaattt

ttteeerrrttt

aaall R D 4545 m

A 4027 m

MR 4634 m

447844784478

W 4505 m

2200 m

2000 m

1800 m

1600 m

0 10km

SWITZERLANDSWITZERLANDSWITZERLAND

(a) (b)

(d)(c)SE NW

Paragneiss

Orthogneiss

Basal fault

Deposit

Rock fall 9.5.91

Rock fall 18.4.91

F104

F103

F102

F116

F101

Fig. 2.12 Randa rock slide. (a) Location map showing position of the Mattertal valley, south-western Switzerland (MR, Monte Rosa;A, Allalinhorn; D, Dom; R, Randa; W, Weisshorn; V, Visp; B, Brig). (b) Records of heights of displacement based on geodeticmeasurements taken in the central part of the sliding rock mass (Source: Götz & Zimmermann 1993). F101, etc., refer to specificmonitoring locations on the slide. (Reproduced with permission of The Japan Landslide Society.) (c) Photograph of the Randa rock slidein 2003. The entrance to the River Vispa bypass tunnel is shown in the foreground. (Photograph courtesy of R.M. Johnson.) (d) Cross-section through the rock face and rock slide deposits. (Source: Quanterra 2003.)

ESC02 9/8/06 11:24 AM Page 61

62 JEFF WARBURTON

as low as possible through pumping and excava-tion of a drainage trench. Eventually a 3.7 kmlong bypass tunnel for the Vispa River was completed (Fig. 2.12c). Following the event, roadand rail connections were rerouted away fromthe site. Costs of works and surveys exceeded110 million Swiss Francs (c. 88 million US$).

The rock slides were produced by structuralweaknesses in the valley-side rock slopes. Reliefjoints parallel to the surface developed up to200 m below the surface. At the base of the slopea steep continuous fracture traversed the slope andthree joint sets divided the face into a large block(Fig. 2.12d). The face was loaded from above bypre-existing instabilities which also increasedwater infiltration into the rock mass, promotingweathering and build-up of porewater pressures.The rock’s mass was estimated to have 7 to 15%void content (Schindler et al. 1993). Prior to theevent small rock-falls had been observed with thefinal 1991 failure occurring during a snowmeltperiod. Jets of water were observed near the basalslip-plane just before and during the vent. Rockslides of this magnitude are relatively rare, andover the past 100 years in the Alps only a dozenevents of similar size have occurred.

Events such as this in Switzerland, althoughinfrequent, are not unusual. One of the mostfamous historic landslide events is the Elm landslide in 1881 (Heim 1882). This developedin unstable, highly deformed sedimentary rocks.Local slate mining was thought to be partlyresponsible for the failure. The mass movementoccurred in three stages. Following failure therock-fall debris hit a rocky ledge on its steepdescent, the debris then cascaded like a waterfallon to lower slopes where the main mass con-tinued downslope towards the village of Elm. A small part of the landslide continued under its own momentum and surged 100 m up theopposite hillslope. The landslide lasted an estim-ated 40 seconds and travelled about 2 km. Thetotal volume is estimated to have been about 10 million m3. Although termed a landslide thedebris did not slide but flowed as a granularmass (sediment-gravity flow) with a dispersivestress developing as a result of grain collisions in the mobile mass. There were 115 fatalities

and local communities were devastated by theevent.

2.3.4 Volcanically triggered mass movements andlong-term sediment delivery

Sediment delivery following major volcanicevents is conditioned by the nature and extent ofsediment deposited from the initial eruption andhow these are reworked over time. An import-ant element is how geomorphological processesact to determine the local dominance of sedimenttransfer processes. Figure 2.13 shows a schematicdiagram of the changing dominance of differentsediment transfer processes with distance fromthe Mount St Helens crater following the 1980eruption (Scott 1988). The section shows a 60 kmtransect from the crater down the Toutle River.Over this distance there is a transition from pyro-clastic surge and avalanche behaviour to debrisflows and finally hyperconcentrated streamflow.In order to assess the impact of a mass movementevent of the like seen at Mount St Helens a usefulway of classifying catastrophic mass movementsis in terms of the runout distances (Siebert 1984).A suitable relationship for describing the beha-viour of different large-scale mass movementstypes is the ratio of vertical drop to travel dis-tance (H/L). Figure 2.14 compares various typesof mass movement using this simple relationship.Although there is considerable scatter in the datait is clear that volcanic landslides have greatertravel distances, with some pyroclastic flows andlahars travelling great distances (H/L < 0.02). Hsu(1975) developed an index to describe the exces-sive travel distance (Le) of large mass movementsthat travel distances greater than the maximumpoint expected by sliding of a body with a 0.62friction coefficient (the value of 0.62 is the frictioncoefficient applicable to sliding of a rigid block).The relationship can be expressed as Le = H/0.62.In terms of Mount St Helens the distances travelled by the pyroclastic flows and lahars wereexcessive, resulting in an extensive tract of deva-station. This special quality of extended runoutmeans that large-scale mass failures in volcanicmountainous terrain should be considered dif-ferently in terms of their sedimentology.

ESC02 9/8/06 11:24 AM Page 62


Along active continental margins earthquakeand volcanic activity can result in greatly enhancedsediment loads. At Mount St Helens in the fourmonths following the eruption the Cowlitz River(a tributary of the Columbia River, see Case StudyFig. 2.2) had a sediment load of 140 × 106 t, an amount that needs to be compared with the

estimated normal annual load of the Columbia of 10 × 106 t. In the years following the eruptionthe Columbia had an annual sediment dischargeof 35 × 106 t (Meade & Parker 1985). Dramaticpost-eruption changes in channel morphology andreworking of volcanic debris resulted in sedimentyields that are two orders of magnitude greater

aaannndddbbbaaasssaaa lll aaavvvaaa lllaaannnccchhheeesss

DDDeeefff lll aaa ttt iii nnnggg

pppyyyrrroooccc lllaaasss ttt iii cccsssuuurrrgggeee

0 10 20 30 40 50 60

Alti

tude

in m

etre

s

0

500

1000

1500

2000

2500

Boundaryapproximate

Phases

RunoutDistalMedialProximal

Channel distance from crater, in kilometres

Alluvialchannel

Bedrock gorge

Expanding alluvialchannel with bedrock

valley sides

Confined bedrock channel

Broad alluvial channel,previously glaciated,

with bedrock valley side

Incisedcanyonson cone

Location ofwhaleback barsDebris flow

Rapid change inflow behaviour

Crater

Co

nfl

ue

nce

Debris flow withrapid textural change

Debris flow progressivelyevolving to hyper-

concentrated streamflow

Toutle River4–10 mm5–20 mm2–5 mm1.3 mm

1.3 mm

0.5–1.0 mm0.5–1.0 mm

Fig. 2.13 Schematic diagram showing the changing dominance of different sediment transfer processes with distance from the craterfollowing the 1980 eruption of Mount St Helens. Variation in dominant grain-size (mm) of the deposits with distance away from thecrater is also shown. (Source: Scott 1988; Courtesy of U.S. Geological Survey.)

H/ L 0.5

H/ L 0.2

H/ L 0.1

H/ L 0.05

H/ L 0.02

Non-volcanic landslides

Volcanic debris avalanches

Pyroclastic flows

Lahars

10

0.1

1

1 50010 100

Ve

rtic

al

dro

p (

km)

Travel distance (km)

Fig. 2.14 Relationship betweentravel distance (L) and vertical drop(H) of large landslides. The value0.62 is the friction coefficientapplicable to sliding of a rigid block.(Reprinted from Journal ofVolcanology and GeothermalResearch, Siebert, L. (1984) Largevolcanic debris avalanches:characteristics of source areas,deposits and associated eruptions,22(304), 163–97.)

ESC02 9/8/06 11:24 AM Page 63

64 JEFF WARBURTON

than pre-eruption levels (Major et al. 2000; Major2003). At Mount St Helens post-1980 sedimentyields declined non-linearly for a decade butincreased abruptly in response to higher thannormal runoff in the late 1990s (Major 2003).Even after 20 years some drainage basins stillhave sediment yields 10 to 100 times greaterthan the pre-eruption levels. Although sedimentsources in small drainage basins tend to stabilizefairly rapidly (Collins & Dunne 1986), sedimentstored in larger river valleys remains active andis unlikely to be stabilized for several decades(Major 2003).

2.4 PROCESSES AND IMPACTS – ANTHROPOGENIC

INFLUENCES

A natural hazard can be defined as ‘A phys-ical event which makes an impact on humanbeings and their environment’ (Alexander 1998).Mountainous environments are highly activegeomorphologically and increasing human useof mountains has led to an increase in natural hazards (Hewitt 2004). In the context of fluvialhazards in mountain environments ‘Mountainrivers become a hazard only when they threatenhuman life or property, by inundation, erosion,sediment deposition or destruction’ (Davies1991). Although a large range of fluvial hazardsoccur (e.g. glacial outburst floods, Alpine debrisflows, etc.) they generally result from extremetemporal and spatial variability in fluvial andhydrological processes. Because of these char-acteristics mountain environments pose uniqueproblems for hazard assessment, prediction andmitigation. More data are required on the fre-quency and magnitude of hazardous events inmountain regions.

This section considers two main examples ofhuman interaction with mountain sediment sys-tems. The first outlines the debate surroundinghuman-induced accelerated soil erosion in a steepupland environment – deforestation in Nepal.The second examines human infrastructure con-struction in an unstable mountain environmentand discusses issues regarding the KarakoramHighway.

2.4.1 Anthropogenic impact on upland sedimentsystems – deforestation in Nepal and the‘Himalayan environmental degradation theory’

The so-called Himalayan environmental degra-dation theory (HEDT) as proposed by Ives &Messerli (1989) neatly illustrates how changesin population structure and pressure (humanimpacts) on a mountain environment can leadto a change in the natural sedimentary system(Fig. 2.15). What is most significant about the theory is the short time-scale over which itappears to have developed and the very largescale of the mountain area it potentially affects.This qualitative model (Fig. 2.15) is deceptivelysimple and can be summarized as follows(Gerrard 1990):• Population growth. Introduction of newhealth care from 1950 has produced rapid population growth. This population explosionhas been amplified by migration from the Indianplain and Nepal with consequent greater de-mands on fuel, construction materials, fodderand agricultural land.• Deforestation. Population expansion resultsin massive deforestation.• Soil erosion and landslides. Deforestation onmarginal and steep mountain slopes leads tocatastrophic soil erosion and landsliding with abreak down in the normal hydrological cycle.• Increased runoff and flooding. The change inslopes has resulted in increased runoff duringthe summer monsoon with increases in flood-ing and sedimentation in the Ganges andBrahamaputra rivers resulting in the extensionof these great river deltas (Fig. 2.15, macro-level).• Positive feedback – accelerated deforestationand greater soil loss. Loss of agricultural land inthe mountains results in greater pressure andmore deforestation and a switch to animal dungas fuel, depriving the land of much needed fertilizer.• Degraded soil structure. Crop yields declineand soil structure is degraded leading to furthersoil erosion and regolith instability.

Although conceptually attractive this simplemodel is fraught with difficulties. These relate

ESC02 9/8/06 11:24 AM Page 64


to the uncertainties attached to gathering hardevidence to test the model. Gerrard (1990) pro-vides a useful review of the main criticisms of thetheory and poses several key questions requiringnew knowledge on the rates of deforestation,the extent of deforestation, the main forms ofsoil erosion and landsliding, the spatial extent of erosion, the nature of the river sediment loadand whether the observed changes are the resultof land-use change or natural variability in thesediment system.

Environmental sedimentology is ideally suitedto answering such problems through better under-standing of the impact of human actions on thesedimentary system. For example, Laban (1978,1979) has shown that very high soil erosion rates

in Nepal can be tolerated within the landscapeand half the landslides are from natural causes.Ironically areas least affected by landsliding arethe heavily populated and terraced agriculturalsystems. In the lesser Himalaya evidence foraccelerated erosion has been found from studiesof the small Nana Kosi watershed where agri-culture and deforestation have increased erosionrates by a factor of five to ten (Rawat & Rawat1994). Such discrepancies in erosion-rate evid-ence from site-specific locations are hard toaccommodate in a general model, however, over-all the evidence generally points to rates of soilerosion that are generally less than predicted andthe sediment system is fairly robust under suchhigh rates of change.

MMM iii ccc rrr ooo lll eee vvv eee lll

MMM eee sss ooo lll eee vvv eee lll

MMM aaa ccc rrr ooo lll eee vvv eee lll

Human impact Natural impact

–

–

–

Precipitation risks(intensity / pattern)


Natural hazards


Floods and sedimentation



Deforestation andland-use intensification

without conservationmeasures

Increasing:- soil erosion- sediment load- runoff

River bed rising dueto water deviation

Local floodsin lowercourse

Fig. 2.15 General summary of the key elements of the Himalayan environmental degradation theory as proposed by Ives & Messerli(1989). Schematic diagram showing the relationship between human-induced and natural processes at the micro-, meso- and macro-levels. (Source: Ives & Messerli 1989; reproduced with permission from The Himalayan Dilemma – Reconciling Developmentand Conservation. Ives, J.D. & Messerli, B. (1989). Routledge, London.)

ESC02 9/8/06 11:24 AM Page 65

66 JEFF WARBURTON

A major part of the problem is assessing therepresentativeness of the available data on ratesof natural and human-induced erosion in spaceand time, especially the relative and absoluteimportance of rare high-magnitude events com-pared with continuous slow mass movements(Ives & Messerli 1989). In general there is no sub-stantive evidence that human activities such asdeforestation, subsistence farming, irrigation androad construction have had a major impact onsoil erosion on the large scale. Locally such pro-cesses may be a problem but at the large scalethe overwhelming importance of geophysical andclimatic controls on the delivery of sediment fromthe rapidly uplifting and eroding mountains tothe Ganges–Brahmaputra plains is paramount(Ives & Messerli 1989). Incorrect diagnosis ofthe cause of the large-scale problem could resultin costly but unproductive changes in land man-agement, the longer term effects of which wouldbe hard to predict.

2.4.2 Road construction in mountainenvironments – the Karakoram Highway

Increased access to mountain areas usually involves the construction of roads and highwaysfor the passage of motor vehicles. This bringswith it a suite of special engineering problems(Fookes et al. 1985). Gerrard (1990) outlinesfive stages in mountain road construction whichusually involve: feasibility and project planning;reconnaissance of road corridors; site invest-igation of the road alignment; construction; andpost-construction maintenance and observation.Routing a road through a steep, unstable moun-tain environment presents major challenges tothe highway engineer at each of these stages. Forexample, if a road is planned across a talus slopethen the alignment should be across the lowerslope because these areas have the lower slopeangles and larger sediment sizes, which providegreater strength during excavation. Further-more rock-fall hazard reduces towards the toeof the slope. Alternatively if a road has to cross amudflow zone the road should be as high on theslope as possible where mudflows are thinnest(Gerrard 1990).

Perhaps the best known and most widelystudied mountain highway is the KarakoramHighway in the Hunza valley, Pakistan (Fig. 2.16).The geographical setting makes road construc-tion and maintenance a continuing problemowing to:1 an overdeepened glaciated valley with verysteep unconsolidated valley slopes;2 an unstable valley floor with very varied terrainconsisting of moraine ridges, outwash fans, gorgesections and sedimentation zones of till, out-wash gravel and debris flow deposits;3 a highly variable hydrological regime withrunoff in excess of 900 mm yr−1 (peak flow > 2000 m3 s−1) and flood history caused by outburst floods from adjacent glaciers or dam-break floods from landslide and debris-flow valley dams;

Fig. 2.16 The Karakoram highway in the Hunza ValleyPakistan. Frequent road maintenance is required to keep the roadopen and free of debris. (Photograph C. Warburton.)

ESC02 9/8/06 11:24 AM Page 66


4 the fact that most of the route of the road tra-verses unstable Holocene and Pleistocene depositsand passes through highly dynamic outwashzones such as the Batura glacier terminus.

A general assessment of the highway under-taken by Jones et al. (1983) showed the relativesignificance (% of length) of different terrain andmaterial types along 129 km of the highway:

alluvial and outwash fan 40.6rock fall and scree 20.7rock 18.3terrace deposits 9.4river 0.4till 10.6

Considering the dynamic nature of the terraintypes and unstable character of the materials itis inevitable that despite best practice in design-ing and building the road, sections of it will beperiodically destroyed during seismic events,floods and debris flows. The only action is torebuild following these natural catastrophes(Fig. 2.16). The value of environmental sediment-ology in designing such a highway is in carefulplanning of the alignment of the road in relationto the geomorphology and terrain types; assess-ment of slope instability types and likely hazards;and siting of local engineering measures to mini-mize risk from specific slope processes.

In more developed mountain countries roadand railway construction is a highly sophisticatedengineering practice involving the constructionof switchbacks, blasting to achieve preferredalignments; elaborate bridges; protective struc-tures from rock fall and avalanches; and tunnels.In the Rhaetian Alps in Switzerland the railwayruns over a distance of 240 km and on this routehas 376 bridges and 76 tunnels (Price 1981). Insome cases tunnels pass completely through amountain such as the 12 km Mont Blanc Tunnel(completed in 1965) between France and Italy.

2.5 MANAGEMENT AND REMEDIATION

Environmental sedimentology is an essentialelement in the management and remediation of

sediment related hazards in mountain regions.This can be illustrated by considering examplesof hazard assessment and mapping techniquesand the role monitoring can play in the remedi-ation of mountain sedimentation events.

2.5.1 Hazard assessment

In mountain environments geomorphological pro-cesses are highly active and the terrain inherentlyunstable (see section 2.1.2). Any move to developsuch areas results in a potential hazard. Moun-tain hazards are on the increase due to increasingdevelopment pressures and recent environmentalchange. Mountain hazard mapping is thereforebecoming an increasingly important componentof regional and local land-use planning. Over thepast few decades understanding of mountain sedi-ment systems has advanced considerably (Owens& Slaymaker 2004), however, the problem ofranking the importance of geomorphologicalprocesses in terms of hazards is still problematicbecause of difficulties in:1 recognizing and understanding causal mechanisms;2 adequately characterizing the time-scale andfrequency over which processes operate;3 accurately mapping the occurrence of suchphenomena.

Slope instability in mountainous terrain is anatural occurrence. In many regions, however,this represents an important hazard that needsto be carefully assessed. In Switzerland, there is a long history of landslide disasters and genuine concern about whether future climaticwarming will lead to an increase in hazards, for example greater frequency of debris flowsfrom the periglacial zone. This is a significantproblem because more than 6% of the countryis affected by hazards that are related to slopeinstability (Raetzo et al. 2002). Examples ofspecific events include the summer debris flowsof 1987 and the Randa rock avalanches of 1991(see section 2.3.3). Following the devastatingevents of the summer of 1987 in Switzerlandrevisions were made to the Federal Flood Protec-tion Law and the Federal Forest Law, which camein to force in 1991 to protect the environment,

ESC02 9/8/06 11:24 AM Page 67

68 JEFF WARBURTON

human lives and property from damage causedby water, mass movements, snow avalanchesand forest fires. It is now a requirement that canton administrative units produce registers ofhazard types and maps of hazard areas for use inland-use planning. Raetzo et al. (2002) describea three-step procedure for the hazard assess-ment of mass movements. This involves hazardidentification, which consists of classification oflandslides, mapping of landslide phenomena anda register of slope instability events. The secondstep involves hazard assessment, which requiresthe determination of the magnitude or intensityof an event over time and the expression of thison a hazard map. The hazard maps are designedto show degree of danger and by convention usecolour coded danger zones (red – high danger;blue – moderate danger; yellow – low danger;and white no danger). These categories are closelyrelated to the intensity and probability of theevent occurring. The third step is risk manage-ment and land-use planning. This utilizes thehazard map, which is the basic document forland-use planning. This feeds directly into thelocal authority planning process. For example,standard colour coded hazard zones indicate the degree of development that can occur, forexample: in red zones construction is prohibited;in blue zones construction is allowed only if certain conditions are met; and in yellow zonesconstruction is permitted without additionalrestrictions.

In Austria a similar scheme for the classifica-tion and mapping of hazardous mountain eventshas been produced for torrents, avalanches andfloods (Aulitzky 1994). This is based on workundertaken by the Austrian Forest TechnicalService for Torrent and Avalanche Control and is legally bound by the 1975 Forest Law. Itwas realized that following a series of majormountain disasters in the 1960s in Tyrol thattwo-thirds of losses were related to some form ofhuman influence. Because most of these hazardsoccur on alluvial cones and fans a formula wasdeveloped which delimits different degrees ofrisk across the fan surface in relation to theamount of debris deposited (G) (Hampel 1980).This is based on an assessment of the gradient of

the alluvial fan (J ) and the mean particle size ofmaterial of the sediment transported (dm):

G(%) =

This simple formula is very sensitive to the measured fan slope (Aulitzky 1994). Additional criteria are used for hazards in torrents (TorrentIndex method) and landslides. This informa-tion is used to produce hazard maps, whichshow hazard zones for events with an estimatedreturn period of less than 150 years. These includea red zone (permanent settlement and trafficprohibited); yellow zone (damage certain – somedevelopment allowed but with restrictions) anda white zone (no recognized hazard).

2.5.2 Monitoring

Hand-in-hand with hazard assessment andmapping goes hazard monitoring. In areas that have been established to be at high risk hazard,monitoring will be undertaken. Usual hazardmonitoring involves direct measurements of the behaviour of a river, slope or glacier, whichcan be relayed in real-time or over short time-scales to an observer who can plan a response,for example activate a warning and evacuationsystem.

The Randa rock slide discussed earlier (sec-tion 2.3.3) provides an excellent example ofhow both short-term (emergency) monitoringand longer term monitoring provide importantinformation. The initial rock slide at Randa inApril 1991 prompted the authorities to install aseismographic and geodetic warning system. As aresult the second event in May was forecast fromfield evidence and geodetic and seismic surveysand the area was successfully evacuated. Prior tothe second slide geodetic displacement measure-ments indicated an accelerated motion of the rockmass (Gotz & Zimmermann 1993) (Fig. 2.12b).Approximately 2.5 × 106 m3 of active block arestill being monitored. The rock mass is still mov-ing towards the south-east at a maximum rate of15 mm yr−1 (Ornstein et al. 2001). The slidingrock mass is bounded by shallow dipping joints

(J% − 55dm1.65)1/(0.42−0.4dm)

3.6

ESC02 9/8/06 11:24 AM Page 68


and sets of steep near-vertical joints. Tensioncracks on the surface behind the slope face tendto open parallel to the steeply dipping naturaljoints. Currently the Swiss Seismological Servicehave installed a network of surface seismographsto monitor local activity. This has proved usefulin identifying fracturing within the underlyingrock mass (Eberhardt et al. 2001). Monitoringwill be extended to included borehole installa-tion of microseismic triaxial transducers aroundthe developing shear plane. The movements are being monitored by geodetic survey and anextensiometer gauge, which will enable the brit-tle fracture development of the rock mass to beassessed. Based on observations of the currentinstability the local authorities have been able toact in advance of another failure, i.e. to realignthe road and railway to avoid the impact ofpotential new landslides.

Case Study 2.3, which focuses on the sedi-ment related problems following the 1980 Mount St Helens eruption, provides additional detailson remediation. It is clear from the Mount St Helens example and the examples just dis-cussed that there are three phases in which envir-onmental sedimentology can play a significantrole:1 hazard assessment and mapping;2 monitoring of mountain sedimentary processes;3 prediction of long-term sediment deliveryfrom disturbed mountain environments anddesign of appropriate remediation measures todeal with the sedimentation problem.

2.6 FUTURE ISSUES

All environmental and economic indicators suggest that over the next few decades moun-tain environments will become more stressedand erosion and environmental degradation willbecome greater problems. This section con-siders how sedimentology can provide usefultools for remediation of geomorphological prob-lems, thus enabling mountain life to continue ina changing environment. The two main examplesinclude: predictions involving the evaluation ofclimate change scenarios and the occurrence of

debris flows, and microscale hydraulic modellingof sedimentary systems.

2.6.1 Considering future climate change scenarios:debris flows in the Alps

Many mountain environments are influenced byglacier and permafrost hazards, which present a direct threat to infrastructure, settlements andhuman life in high-altitude and high-latitudemountain areas. Global warming is affecting thethermal stability of surface and subsurface ice,and glacier and permafrost equilibrium limitsare shifting in response to this general trend (see Chapter 1). The situation is made worse asdevelopment pressures in mountain areas areforcing human settlement and activities fur-ther into the mountain cryospheric zone. Forexample, under the prediction of a warming scenario for the Alps the following are all likelyconsequences (Zimmermann & Haeberli 1992;Haeberli 1995):1 greater frequency of outburst floods andglacier hazards;2 permafrost degradation and slope instability– increased debris flow activity;3 decline in material strength of foundations ofhigh mountain buildings;4 damage to reservoir systems (damage to infra-structure) and increased sedimentation rates(reduced storage capacity).

More generally the impact of such warming isa range of glacier-related hazards that include:glacier lake outburst floods; rock and ice ava-lanches; deep seated instability; and retreat ofglacier and permafrost limits. Many of thesehave important impacts on the sedimentology ofmountain environments, resulting in enhancedrates of debris supply, extensive flood deposits andlarge-scale mass movements. Examples includethe increased frequency of rock avalanches inrelation to permafrost degradation in glacierenvironments: the Brenva Glacier, Mont Blancin 1997 (Bottino et al. 2002; Haeberli et al. 2002)and the September 2002 Kolka–Karmadon rock and ice avalanches and mudflows in theCaucasus Mountains of Russia (Kääb 2002;Kääb et al. 2003).

ESC02 9/8/06 11:24 AM Page 69

70 JEFF WARBURTON

The Kolka–Karmadon rock avalanche beganon the evening of 20 September 2002 high on thepeak of Dzimarai-khokh. Several million cubicmetres of ice and debris fell on to the Kolkaglacier tongue, shearing off the front of the glacierbefore crossing the Maili glacier. Travelling ataround 100 km h−1 the avalanche picked up lateral moraines and valley-bottom sedimentsbefore running out and halting approximately18 km from the source, where approximately80 × 106 m3 of ice and debris were deposited andafter which a mudflow approximately 300 mwide continued for another 15 km down the con-fined valley. A total of 120 people were killed inthe catastrophe and since the event the depositscontinue to influence sedimentary processes inthe valley system. Rivers were dammed by theavalanche deposits, which caused the forma-tion of lakes estimated to be over 10 × 106 m3

in volume. These posed a considerable danger to settlements down-valley owing to the risk ofdam-break flooding.

Because of the remote nature of the environ-ments in which these high-mountain processesoperate, rapid assessment of glacier hazards is needed through the use of remotely senseddata. Rapid imagery from ASTER (AdvancedSpaceborne Thermal Emission and ReflectionRadiometer) provides 15 m resolution imagery,which is updated regularly every 16 days but thiscan be shortened to 2 days where needs demand(Kääb et al. 2002). From these data digital elevation models (DEMs) can be generated,which are essential for assessing the dynamicchanges in topography that characterize theserapid sedimentary events.

The above example illustrates how recentenvironmental change is having an impact onmountain hazards. It should be recognized, how-ever, that the incidence of mountain hazards alsovaries over much longer cycles. For example,Soldati et al. (2004) examine the relationshipbetween the temporal occurrence of landslidesand climatic changes in the Italian Dolomitessince the last glacial stage (Würm). Increases in landslide activity were identified correspond-ing to the boundaries between the Late-glacialand Holocene and the Atlantic and Sub-boreal.

Landslide type and cause also differed overtime. Following retreat of the glaciers after the Last Glacial Maximum there followed aphase of increased slope instability (11,500 to8500 cal. yr BP) involving large translationalbedrock slides in the dolomitic slopes and complex movements (both rotational slides andflows) affecting the underlying pelitic forma-tions. A second phase of enhanced activity wasalso identified around 5800–200 cal. yr BP, whenslope processes were dominated by mainly rota-tional slides and flows. The mechanisms for suchactivity vary. In the earlier slides high ground-water levels related to increased precipitationand/or permafrost melt were probably import-ant. In the second phase, many of the slopemovements were reactivations of earlier land-slides and probably were triggered by increasedprecipitation. Studies of this kind are valuablewhen considering potential impacts of futureclimate change because although patterns of historical landscape instability can be linked toclimate, and to some extent these fit with gen-eral patterns of slope movements across Europe,non-climatic factors (e.g. geological–structuralfactors and changing human land-use) also havea role to play and should not be excluded fromthe analysis.

2.6.2 Microscale modelling

Understanding sedimentation problems inmountain environments is difficult because prob-lems evolve over several decades or centuries andthe main governing processes operate at scalesthat cannot be measured easily. Microscale model-ling provides a valuable tool for the environ-mental sedimentologist to investigate large-scalesedimentation processes in mountain-valley sedi-ment systems. Microscale loose-bed hydraulicmodels are very small-scale and conventionallyhave been used to evaluate channel designs on large river systems (Gaines & Maynord2001). Typically model studies are carried outto determine channel response to engineeringstructures and identify overall flow patternswithin the fluvial system. A basic model consistsof a hydraulic flume, a rigid channel insert to

ESC02 9/8/06 11:24 AM Page 70


define the valley-system boundaries and a loosebed of natural or synthetic bed material. Typicalmodel scales are in the order of 1:10−3 or 10−4.At this horizontal scale microscale models dif-fer from traditional loose bed hydraulic modelsbecause in order to ensure that sediment trans-port occurs in the model, water depths are exaggerated and as a consequence vertical scaledistortion is inherent. Also shallow depths leadto surface tension effects and laminar flows,which do not occur in the real river. The successof microscale models, however, is not judged on the basis of hydraulic similarity but more soon how well overall the model reproduces thegross features of the prototype river morpho-logy (Gaines & Maynord 2001). In this sensethe approach is useful in that it provides an initial assessment method for rapidly simulatingriver behaviour at large scales.

The technique has been used by Davies et al.(2003) to examine anthropogenic aggradationof the Waiho River, Westland, New Zealand.Long-term aggradation of the Waiho River alluvial fan over much of the past 100 years has raised the fanhead to an unprecedented highlevel. This now poses an unacceptable flood risk to the adjacent village of Franz Josef Glacierand the main State Highway 6, which is a keytourist corridor. The behaviour of this fan haspuzzled geomorphologists for some time andhas led to a number of hypotheses regarding the evolution of the fan system in relation toriver control measures (Davies 1997; Davies & McSaveney 2001). The only suitable way of testing these ideas, however, was to construct a model capable of simulating river behaviourfor a variety of imposed boundary conditions. A 1:3333 microscale physical hydraulic modelwas used to study the problem (Fig. 2.17a). Inthe model an alluvial fan was generated andallowed to develop to equilibrium with steadyinputs of water and sediment. The fan was later-ally constrained by rigid boundaries which weregeometrically similar to the natural unrestrictedWaiho River. The boundaries were then reset toreflect the presence of stopbanks (river controlstructures) and the fan allowed to evolve under thesame water and sediment feed rates. The model

fanhead aggraded in a similar spatial pattern to that observed in the real river. Figure 2.17bshows a comparison between relative changes in bed level in the model and those measured in the Waiho River. The correspondence in thespatial pattern of aggradation was very similar.This implies that the reduction of the flow areaat the fanhead caused by the river stopbanks is sufficient to cause the observed aggradation in the Waiho.

This example clearly demonstrates the advant-ages of using microscale modelling for under-standing large-scale sedimentation problems inmountain environments where rates of sedi-mentation are relatively rapid and topographyexerts a major control on the spatial patterns of cut and fill. Although there are constraints inusing this type of model (e.g. hydraulic condi-tions cannot be truly scaled) a formal physicalmodel would require a planform of 50 m by 50 m which is beyond the scope of most physicalmodel investigations.

2.7 CONCLUDING COMMENTS

The aim of this chapter has been to show howthe functions of active sedimentary processesare altered by human actions and recent climate/environmental change in mountain environ-ments. It has been demonstrated that mountainenvironments are characterized by high relief,steep slopes and local climates that vary overaltitude. These environmental constraints createa distinctive set of mountain sedimentologicalprocesses. Figure 2.18 is a simple conceptualmodel that summarizes the relationship betweenmountain sedimentation zones and dominantprocess regimes. The diagram distinguishes sedi-mentation zones in terms of a sediment-cascadecontinuum which links slopes and valley-floorsediment systems along a gradient of downslope/valley sediment fining. Dominant process regimesmap on to the sedimentation zones but theseoverlap to differing degrees dependent on theextent of their spatial influences (broad boxesshow the main zone of operation and the arrowsshow the maximum extent). The large overlap

ESC02 9/8/06 11:24 AM Page 71

72 JEFF WARBURTON

300

-100

100

50

0

-50

0 12080604020

250

200

150

100

(a)

(b)

Model1985–1993

Relative distance between SH6 bridgeand Waiho Loop (%)

Rel

ativ

e in

crea

se in

bed

ele

vatio

n (%

)

Fig. 2.17 Microscale model of the Waiho River fan, New Zealand. (a) Laboratory model showing unconstrained fan development.(Photograph courtesy of T.R. Davies.) (b) Comparison between relative changes in bed level in the model and those measured in the Waiho River. Bed-level changes are non-dimensionalized by scaling the variance in each data set to 1.0. (From Anthropogenicaggradation of the Waiho River, Westland, New Zealand: microscale modelling. Davies, T.R., McSaveney, M.J. & Clarkson, P.J. (2003) Earth Surface Processes and Landforms, 28, 209–18. © John Wiley & Sons Limited. Reproduced with permission.)

ESC02 9/8/06 11:24 AM Page 72


between sedimentation zones and process regimesin mountain environments leads to extremevariability in space and time of mountain pro-cesses and makes understanding of such sedimentsystems problematic (Gerrard 1990), with theinevitable conclusion that sediment erosion anddeposition in mountain environments is highlyheteorogeneous (Butler et al. 2003).

Nevertheless, understanding how these varioussedimentological processes operate in mountainenvironments has been considerably aided in the last two decades by a series of methodo-

logical advances. In particular it is worth notingthe importance of catchment-based sedimentbudget studies (Rapp 1960; Barsch & Caine1984; Trustrum et al. 1999; Slaymaker et al.2003). Looking to the future, similar advancesin understanding are likely to be made through:the application of geophysical techniques forestimating sediment storage (Schrott et al. 2003);use of cosmogenic dating to better understandtime-scales of erosion and sediment residence(Kirchner et al. 2001); use of remote sensing inhazard monitoring and mapping (Kääb et al.

Mountain

sedimentation

zone

Dominant

process

regime

Coa

rse

Fine

Slo

pes

Lake

sed

imen

tatio

n

Gul

ly s

yste

ms

Allu

vial

fans

and

con

es

Valle

y flo

or

Str

eam

cha

nnel

Slope mass movements

Channelized debris flow

Flood flows

Glaciation

S e d i m e n t c a s c a d e

Fig. 2.18 Simple conceptual model relating mountain sedimentation zones to dominant process regimes.

ESC02 9/8/06 11:24 AM Page 73

74 JEFF WARBURTON

2003); and use of modelling to predict futurechanges in mountain sediment systems (Bottinoet al. 2002; Davies et al. 2003).

Finally although this chapter has been aboutmountain environments, it is important to stressthe links between mountain regions and low-lands. Erosion in the uplands has a profoundimpact on downstream sedimentation (e.g.

Brahmaputra and Ganges rivers draining theHimalaya) and at the junction between moun-tain and lowland environments sedimentolog-ical processes are often most active (e.g. alluvialfan aggradation, Waiho River, Westland, NewZealand). Sediment processes in lowland riversystems are explored in Chapter 3.

ESC02 9/8/06 11:24 AM Page 74

3.1 INTRODUCTION

Rivers are arteries for the transport and storageof physically and chemically weathered materialfrom continents, through estuaries, and ulti-mately to oceans; they thus play a major role inthe Earth’s biogeochemical cycling of materials,and influence the Earth’s climate. Rivers arefound in every corner of the world, and in everyclimatic zone. Both natural processes and anthro-pogenic exploitation of natural resources fromprehistoric times through to the present day have,however, had a very significant impact on thehydrology, sediment regime and contaminationof rivers world-wide. During the Quaternary, forexample, river systems have experienced con-siderable natural hydrological variability due todeglaciation and sea-level change, and humanactivity has left very few rivers in a pristine state,except perhaps for some in Canada, Amazonia,the Congo Basin and Siberia (Meybeck 2003).

This chapter describes the nature of fluvialsedimentary environments, types and sources ofsediments in these environments, processes andimpacts of natural and anthropogenic disturbanceevents on river sediment fluxes and effects, man-agement of fluvial sedimentary environments and,finally, issues concerning fluvial environmentsthat need to be addressed in the future. The chap-ter draws on the voluminous fluvial geomor-phological, sedimentological, environmental andarchaeological literature. For further informa-tion, the reader is referred to this literature, andto the many excellent books and book chapterson fluvial sedimentary environments, and thebackground information and references containedtherein (Richards 1982; Brown 1997; Thorne

3Fluvial environments

Karen Hudson-Edwards

et al. 1997; Benito et al. 1998; Knighton 1998;Leeder 1999; Miller & Gupta 1999; Bridge2003). This chapter summarizes, updates andbuilds on this work, and, specifically, highlightsthe ever increasing effects of anthropogenicactivity on sedimentation in rivers.

3.1.1 Definition and classification of fluvialenvironments, and relevance to environmentalsedimentology

The word ‘fluvial’ pertains to a stream or river,the existence, growing or living in or about astream or river, or something that is produced bya river (Bates & Jackson 1980). Fluvial environ-ments occur on every continent on Earth and inevery climatic zone (Fig. 3.1), and thus are oftenclassified according to climate, with arid, semi-arid, temperate (cool), temperate (Mediterranean)and tropical types known (e.g. Jansen & Painter1974). Rivers are also classified by their flowregimes, with perennial (flowing every year,throughout the year), intermittent (flow for onlypart of a year, every year, usually during or after a wet season) and ephemeral (occasionalflow) types defined. Intermittent and ephemeralrivers often occur in Mediterranean or tropicalenvironments, whereas perennial rivers occur intemperate regions.

‘Fluvial sediments’ are those sediments thatconsist of material transported by, suspendedin, or laid down by a stream (Bates & Jackson1980). They are important sources of nutrients,contaminants and other solid materials todownstream fluvial, estuarine and coastal envir-onments. Millions of tonnes of sediment are trans-ported annually to oceans (Table 3.1), with the

ESC03 9/7/06 3:16 PM Page 75

76 KAREN HUDSON-EDWARDS

amounts determined by climate (precipitation),topography, vegetation cover, land-use, and thesusceptibility of the underlying rocks, soils orother unconsolidated materials to physical, chem-ical and biological weathering.

A river system is, by definition, the system ofconnected river channels in a drainage (catchment)basin (Bridge 2003). The system contains a largenumber of features, but the two most importantare its channels and floodplains (Fig. 3.2). In the upland, mountainous parts of river basins,relief is relatively high, channels are incised intobedrock or alluvium, and floodplains tend to be narrow. For a more detailed discussion ofsedimentology in mountainous environments,the reader is directed to Chapter 2. In the middleto lowland portions of river systems, relief islower, valleys are broad and larger-scale flood-plains are developed. From a sedimentological

perspective, rivers are often classified accordingto their planform channel geometries, with fourmajor types traditionally defined: (i) straight, (ii) meandering, (iii) braided and (iv) anastomos-ing (Fig. 3.3). This classification has been seen asflawed, mainly because the classes overlap. As aresult, authors such as Rust (1978) and Bristow(1996) have suggested that these types shouldnot be viewed as separate entities but, rather, asend members of a cyclical continuum (Fig. 3.4).Modifications from one end member to anothercan be a result of changes in sinuosity (riverlength divided by valley length) and the amountof braiding (channel length divided by valleylength or island length divided by river length).For example, straight rivers can metamorphoseinto meandering rivers by an increase in sinuos-ity, and meandering rivers can change intobraided rivers through the development of bars

12

43

56

920

18

17

161514

13

12

10 11

21

24 25 26 27

28

22 29

23

30

311 Yukon2 Mackenzie3 St Lawrence4 Mississippi5 Colorado6 Rio Grande7 Orinoco8 Amazon9 Parana

12 Senegal13 Volta14 Niger

10 Rhine11 Danube

15 Lake Chad16 Nile17 Congo18 Zambezi19 Okavango20 Orange

24 Volga25 Ob26 Yeni-sei27 Lena28 Amur

29 Hwang-ho30 Yangtse31 Mekong

21 Tigris- Euphrates22 Indus23 Ganges19

7

8

PolarHumidTropicalDry

CLIMATE ZONES

Fig. 3.1 Climatic zones and the world’slargest river basins. (Based on Newson1992, fig. 4.1. Climate zones from The Times Atlas of the World 1983.)

Table 3.1 Transport of sedimentary material from continents to oceans. (Based on Knighton (1998), with data from Degens et al.(1991), Meybeck (1979), Milliman & Meade (1983) and Walling (1987).)

Continent Land area Mean annual runoff Total annual suspended (106 km2) (103 km3) sediment load (106 t yr−1)

Africa 15.3 3.4 530Asia 28.1 12.2 6433Europe 4.6 2.8 230North and Central America 17.5 7.8 1462Oceania and Pacific Islands 5.2 2.4 3062South America 17.9 11.0 1788

ESC03 9/7/06 3:16 PM Page 76

FLUVIAL ENVIRONMENTS 77

or islands. Some fluvial sedimentologists groupthe straight and meandering types together as‘single channels’, and the braided and anastomos-ing types together as ‘multiple channels’.

The planform geometry of rivers is controlledby climate, tectonics and land-use, and more spe-

cifically by the rate of flow, flood-related sedi-mentary processes and the amount of sedimentdischarge (Schumm 1968; Knighton 1998). Littlechannel change occurs during low flows and lowamounts of sediment transport. By contrast, theincreases in discharge and sediment transport

Drainage divide separates drainage basin from others

Mountain belt with steep valley walls, bedrock channels,alluvial channels, andnarrow floodplains(see also Chapter 2)

Alluvial fans as main channelsemerge from mountain belt

Lake or sea

DeltaDistributary

Main channelTributary

Broad, low-relief valleyswith alluvial channels andfloodplains

Fig. 3.2 Plan geometry of a hypothetical river system. (Based on Bridge 2003, fig. 1.1.)

(a) (b)

Fig. 3.3 Photographs of river channel types. (a) LANDSAT image of the Brahmaputra River in Bangladesh. This is a very large sand-bedbraided river, flowing from north to south down the photograph. On the floodplain on either side are sinuous rivers that act asdistributary and tributary channels draining the floodplain. (b) LANDSAT image of the anastomosed Meghna River in Bangladesh,where the sinuous river channels divide and rejoin around areas of floodplain and are separated for more than one meander wavelength.Width of photographs is approximately 60 km. (Photographs donated by A. Carter.)

ESC03 9/7/06 3:16 PM Page 77


that occur during floods can cause moderate tosometimes drastic channel change (Schumm &Lichty 1963). Attempts have been made to modelchannel changes that occur in response to changesin water and sediment discharge. Schumm (1969),for example, suggests that a decrease in stream-flow and concomitant increase in sediment supplyshould produce a decrease in channel sinuosityand depth, and either an increase or decrease in

Fig. 3.4 Continuum between principal planform types of rivers,illustrating the diversity of river channel types. Between thestraight, meandering, anastomosed and braided types ofplanform are rivers that exhibit characteristics of at least two ofthese end-members. (After Bristow 1996.)

channel width. Channel adjustment, however, is not easy to model, because it depends on several factors including slope, and the pre-existing channel geometry, planform and bedconfiguration. Further examples of the typesand causes of channel change are discussed laterin this chapter.

Fluvial (also known as alluvial) deposits arecharacterized by a huge variety of bed forms andsedimentary structures, which are well-describedin Collinson (1986) and Bridge (2003, chapters 4and 5). These include longitudinal, bank-attached,linguoid, side, diagonal, point and transverse bars,islands, riffles, ripples, pebble/cobble clusters,sand ribbons, cross-stratification, dunes, sandflats, plane beds and chutes-and-pools (largely inchannels), lateral deposits (in channel margins),alluvial fans (in piedmont areas), crevasse splays,levees and vertical accretion deposits (in flood-plains) (Fig. 3.5). These features are used bygeomorphologists to deduce the history of riversedimentation, to evaluate contamination andto record flood histories.

No study or book on environmental sediment-ology would be complete without considerationof fluvial sedimentary environments. Because allhumans essentially reside in river basins, theyhave, since pre-history, had a significant impact onthem and the functioning of their sedimentaryprocesses. In many cases, these impacts outstripthose of their natural counterparts. In othercases, natural (section 3.3) and anthropogenicprocesses (section 3.4) act in tandem to producelarge-scale modifications to fluvial sediment trans-port and deposition.

Anastomosed

Braided

Straight

Meandering

Abandonedchannel-fill

Channel lag

Crevasse-splay

Floodplain

Point-bardeposit

Levee

Earlier deposit

Thalw

eg

Decreasinggrain size

Cross-bedding

Cross-lamination

Flat-bedding

Fig. 3.5 Classic point-bar model for a meanderingstream, showing various types of fluvial sedimentarydeposit. (After Allen 1964, 1970; Collinson 1986.)

ESC03 9/7/06 3:16 PM Page 78


3.1.2 Significance of fluvial sedimentaryenvironments

In effect, many of the issues and problems relatedto river basin disturbance are in turn related todifferences between the Developed world, withits largely temperate rivers, and the Develop-ing world, with its largely arid, semi-arid andtropical rivers. Issues for river development andmanagement can be quite different for these twoworlds. Many of the largest river basins in theworld occur in the arid and semi-arid Develop-ing world (Fig. 3.1), and here the major problemsinclude drought, irrigation, poverty, salinization,pesticide contamination and desertification, all ofwhich have an impact on river systems (Newson1992; see also Chapter 5). By contrast, theDeveloped world is coming to terms with man-aging and restoring river systems that have beenunder pressure from impacts such as contamina-tion. Both the Developed and Developing worldswill have, increasingly, to deal with the issue ofclimate change related to global warming.

River systems are not country-selective: theycross international boundaries. This fact raisesimportant issues relating to sediment transportand deposition in rivers, especially with respectto contamination, erosion and flooding. Newson(1992) has called for workers in all related dis-ciplines to consider the impact of contemporarydecisions about river management in both theshort- and long-term, consider basins as integratedwater, sediment and contaminant entities, and forscience to inform, rather than solve, problems ofriver basin (including sediment) management.

3.2 SEDIMENT SOURCES AND ACCUMULATION PROCESSES

3.2.1 Characteristics and provenance of fluvial sediments

3.2.1.1 Characteristics of fluvial sediment

Transported fluvial sediment can be clas-sified as either bed load, the coarse sediment (> 0.0625 mm) carried along the river bot-tom, or suspended load, the finer sediment

(< 0.0625 mm) moved in suspension. The sus-pended load forms most (generally > 90%) of a river’s sediment and, because its settling velo-cities are generally low, is transported at thesame speed as the river’s flow. During floods,the size of the suspended load can increase dra-matically due to increased stream power. Thesuspended load is supplied from the physicalerosion of river banks and surface materials,and resuspension of fine channel-bed material.

Fluvial sediment also can be classified accord-ing to either its physical or chemical charac-teristics (Goudie 1990). Physically, sediment isclassified mainly according to its size and particleshape. Estimating the relative proportions of the different size fractions in a river sedimentaryunit is important because the size is related to thesource and abrasion during transport, and sizealso plays a major role in the distribution of con-taminants. Particle shape is also determined to a large part by the abrasion and corrosion thatoccurs during transport, but also depends on the original weathering that took place to formthe grain, and on post-depositional chemicalprocesses such as dissolution and precipitation.Particle shape is described in three ways: (i) form,(ii) sphericity and (iii) roundness (see Chapter 1).River gravels, for example, often have high tomedium sphericity and are subrounded. In gen-eral, the further a river sediment particle travelsand the more mechanical abrasion it undergoes,the more well-rounded, higher in sphericity andfiner-grained it will become. Rounding occursmore readily in cobbles and pebbles than in sandsor smaller-sized particles.

The physical and chemical characteristics offluvial sediments are determined by the sedimentprovenance, that is, the origin (source) of the sedi-ment and the physical, chemical and biologicalprocesses that have operated on it during trans-port to the site of deposition in the river system.Most fluvial sediment is composed of geogenicmaterials such as bedrock, soil and vegetation thatare released to river systems through both natural(e.g. erosion, forest fire, volcanic eruption) andanthropogenic activities (damming, deforesta-tion, mining), many of which are discussed insection 3.3. The variability of these geogenic

ESC03 9/7/06 3:16 PM Page 79


materials, in turn, depends on the underlying riverbasin geology, the nature of soil-forming pro-cesses (climate, typography, organisms) and typesof vegetation present. For example, Schell et al.(2000) were able to show that a combination of ore, mine waste, mine-derived alluvium andbedrock, derived from natural weathering, con-tributed to the fluvial sediment load in the mining-affected Rio Tinto basin, Spain (Fig. 3.6a), throughthe use of sediment fingerprinting techniques(see section 3.2.1.2).

Many anthropogenic activities affect the com-position of river sediments (Table 3.2). Theseactivities often add ‘excess’ amounts of sub-stances, or ‘contaminants’, to river systems. Theconcept of ‘contaminant’ has been defined inChapter 1. Contaminant concentrations of sedi-ments are controlled by their abundance in therocks and soils of the basin (the ‘background’concentration) and by the addition, throughanthropogenic activities, of excess amounts of theelement or compound (the contamination). In

the case of artificial compounds (e.g. pesticides,polychlorinated biphenyls (PCBs), some radio-nuclides) the background figure is generally zero, and contamination assessment is relativelystraightforward. Pre-industrial historical valuesneed to be established for elements with bothnatural and anthropogenic sources, such as metals. In rare cases, this can be accomplishedthrough analysis of archived samples, but inmost cases it is determined through the analysisof sediments accumulated through time.

Table 3.2 lists the main types of sediment-boundcontaminants found in river systems. These aregrouped into metals and metalloids, inorganiccompounds, nutrients, organic compounds andradionuclides. The partitioning of each of thesecontaminants will be different, depending on theirgeochemical properties and affinities for sorptiononto the different mineral and organic phasespresent in suspended river sediment. This, inturn, will affect the transport and storage of con-taminants in river channels and floodplains.

Ore/waste

Historical alluvium

Volcanics

Quaternary/blue marls/limestone

0 20 40 60 80 100

60

59

58

57

56

Sediment source contribution (%)

Eff

icie

ncy

(%)

Niebla

Tweed Teviot Ettrick0

20

40

60

80

100

Rel

ativ

e co

ntrib

utio

n (%

)

Woodland topsoilPasture/moorland topsoilCultivated topsoilChannel bank/subsoil

(a)

(b)

Fig. 3.6 Examples of sediment fingerprintingstudies. (a) Results of fingerprinting forsediments at Niebla, Río Tinto, Spain. Efficiency(%) is a measure of the effectiveness of themixing model. The optimal efficiency is shown at the top, and the four nearest efficiency stepsare also indicated. (Based on Schell et al. 2000,fig. 26.4b.) (b) Weighted contributions ofwoodland topsoil, pasture/moorland topsoil,cultivated topsoil and channel bank/subsoil inthe suspended sediment load of Rivers Tweed,Teviot and Ettrick, UK, from January 1996 to February 1997. (From Owens et al. 2000, fig. 15.3a.)

ESC03 9/7/06 3:16 PM Page 80


Sediment-bound contaminants enter riversystems either from point (e.g. tailings effluentand other mine discharges, sewage discharges,spillages) or diffuse sources (e.g. remobilizationof contaminated alluvium, agricultural runoff)(Macklin 1996; Walker et al. 1999). Point sourceseither operate only once (e.g. the Aznacóllar minetailings dam failure in south-west Spain, April1998) or repeatedly (e.g. sewage discharges; regular tailings effluent discharge into the RíoPilcomayo, Bolivia; Hudson-Edwards et al. 2001).Although diffuse contaminants enter rivers con-tinuously, growing evidence suggests that theyare mainly mobilized during storm events. Forexample, Kratzer (1999) suggested that runofffrom infrequent winter storms would continue todeliver significant quantities of sediment-boundorganochlorine pesticides to the San JoaquinRiver, California, even if irrigation-induced sedi-ment transport was reduced. Muller & Wessels(1999) showed that approximately one-third ofthe annual inputs of total organic C, N, Cu, Pband Zn to the River Odra, Poland, were releasedinto the river during a major flood in 1997.

3.2.1.2 Sediment provenance and source materialfingerprints

Analysis of fluvial sediment provenance can yieldimportant information on the types of sources(e.g. topsoil or channel bank materials), relativesource contributions, and on patterns of erosion,

transfer and storage. Although it is possible todirectly determine fluvial sediment provenance,indirect methods have been favoured in recentdecades (cf. Peart & Walling 1986). One of themain types of indirect approaches is that of sediment source ‘fingerprinting’, which has beenused to great effect in studies of suspended andalluvial sediment provenance, over a wide rangeof time-scales ranging from the event-level torecent historic and Holocene. Owens et al. (2000),for example, traced contemporary sources ofalluvium in the Rivers Tweed, Teviot and Ettrick,UK, using composite fingerprints (metallic ele-ment geochemistry, radionuclides, organic C, N and P) and a numerical mixing model, andshowed that most of the sediment was derivedfrom pasture/moorland topsoil and channel bankerosion (Fig. 3.6b). By contrast, Passmore &Macklin (1994) demonstrated that pre-eighteenthcentury alluviation in the River Tyne, UK, wasrelated to deforestation and agricultural develop-ment during late prehistoric times. Walling et al.(2002) used 137Cs measurements and sedimentsource fingerprinting to develop sediment budgetsthat showed sediment tile drains to transferbetween 30 and 60% of sediment from two low-land, agricultural basins in the UK.

The scientific basis of sediment fingerprintingis that the properties of the fluvial sediment arecompared with the same properties of all poten-tial source materials (e.g. Walling et al. 1979).‘Tracers’, which exhibit conservative behaviour

Table 3.2 Types and sources of natural and anthropogenic inputs to fluvial sediments.

Category

Bedrock, soil, vegetation

Metallic and metalloid elements (Sb, As, Cd, Cu, Co, Cr, Pb, Hg, Ag, Tl, Sn, Zn)

Inorganic compounds (SO4, PO4)

Nutrients (C, N, P)

Organic compounds (pesticides, herbicides, petroleum hydrocarbons, viruses, bacteria)Radionuclides (137Cs, 129I, 239Pu, 230Th)

Sources

Physical weathering through natural erosion, forest fires, tectonicuplift, deforestation, agriculture, mining, damming and otherengineering work, urbanizationNatural sources, industry, mining and processing, acid minedrainage, sewage treatment, agriculture, vehicle emissions androad runoff, coal combustion, atmospheric falloutNatural sources, mining and processing, industry, acid minedrainage, acid deposition, agricultureAgricultural and urban land runoff (fertilizers), wastewater fromsewage treatmentAgriculture, industrial processes that produce dioxins, sewage,landfillsNuclear power industry, military, natural sources

ESC03 9/7/06 3:16 PM Page 81


during erosion and transport and which can, on a statistical basis, effectively differentiatebetween the various sources, are used for sourceapportionment (Foster & Walling 1994). Oncesuitable tracers are established, their values, or‘fingerprints’ for potential source materials arecompared with the corresponding tracer valuesfor the ‘unknown’ sediment samples. It has beenincreasingly recognized that no single diagnostictracer is capable of discriminating a range ofsources, so multiple tracer studies using severalfingerprints are now normally used (e.g. Wallinget al. 1993). Over the past several decades, awide range of physical and chemical fingerprint-ing techniques have been developed and used to great effect in determining the provenance offine-grained suspended and alluvial sediment.The main techniques used are summarized inTable 3.3.

3.2.2 Controls on sediment supply, transport andaccumulation

The natural sediment load in rivers is suppliedby physical and biochemical weathering, whichare in turn controlled by (i) the geochemistry,mineralogy and structure of the eroding rocks andsoils, (ii) precipitation, (iii) temperature and (iv)vegetation cover (Bridge 2003). Ultimately, these

are all controlled by climate, topography andland-use. Areas with steep slopes, cold temper-atures and high amounts of rainfall (such as theHimalayan mountain range) have high sedimentsupplies, whereas lowland areas with warmertemperatures tend to have relatively lower sedi-ments loads (which are dominated by clay-richmaterials) and higher dissolved element loads.

Sediment transport in rivers is mainly unidir-ectional, following the dominant flow path, butit can be influenced by turbulence and secondaryflows (Sear et al. 2000). The transport of indi-vidual sediment particles occurs by entrainment,transport and deposition (Hassan & Church1992; Sear et al. 2000). A large number of fac-tors influence sediment transport and depositionin river systems. Among these are climate andseason (Boyden et al. 1979; Macklin 1996), sitegeomorphology (Macklin 1996), streamflowand suspended sediment concentrations (Yeats& Bewers 1982; Kratzer 1999), water depth,flow dynamics, bed surface structure and grainsize and texture (Eyre & McConchie 1993;Macklin 1996). Anthropogenic disturbances,such as channelization, damming and dredgingfor navigation, also play a role in sedimenttransport and deposition.

Although sediment can be mobilized and trans-ported during normal flow conditions, floods

Table 3.3 Fingerprinting techniques used in fluvial sediment provenance studies, and selected references.

Fingerprinting technique

Radionuclides (e.g. 7Be, 137Cs, 210Pb, 226Ra)Mineral magnetism

Sediment mineralogy

Heavy mineralogy

Clay mineralogy

Major and trace element geochemistry

Grain size and shape

Method

Measurement of concentration of radionuclide andcomparison with a background concentrationMineral magnetic parameters (e.g. susceptibility,frequency-dependent susceptibility, isothermalremanence magnetization) measured, statistical analysisMineral abundances determined by X-ray diffraction(XRD)Heavy minerals separated, abundances determined bypoint countingClay minerals separated; abundances determined by XRD and analysis of resultsDetermination of geochemical composition and statistical analysis (e.g. multivariate, principalcomponents analysis)Macro- or microscopic determination of grain size,statistical analysis

Reference

Peart & Walling 1986

Walling et al. 1979

Woodward et al. 1992

Singh et al. 2004

Wood 1978

Lewin & Wolfenden 1978;Passmore & Macklin 1994;Collins et al. 1997Knox 1987

ESC03 9/7/06 3:16 PM Page 82


(high flow stages) play a major role in erodingand depositing fluvial sediment, and modify-ing river channels and floodplains (Knighton1998). During large floods bank erosion up-stream can result in downstream overbankdeposition of fine-grained sediments as thinsheets over surfaces of previously-formed ter-races, or of coarser-grained gravel and sand as sheets, lobes and splays. During relativelymoderate floods that do not exceed bank fullconditions, coarse to fine material is transporteddownstream and eventually deposited as bars or overbank units on lower elevation terraces(Fig. 3.7). The accumulation and preservation ofsediments within river systems also depends onthe factors outlined in the previous paragraph,and particularly on the availability of space andthe deposition and erosion rates (Schumm &Lichty 1963). Deposition on floodplains dependson flood characteristics such as frequency, dura-

tion and suspended sediment concentrations(James 1985). The length of time that sedimentremains stored in a part of the river system isvariable, ranging from 1 to 104 years (Fig. 3.8).

Bridge (2003) has pointed out that all dis-charges that result in sediment transport canaffect river channel geometry. He defines dynamicequilibrium as adjustment of the channel to relatively consistent flood discharges over decadaltime periods, and disequilibrium as a major andregional channel adjustment to a significant event(or events), such as an extreme flood or anthro-pogenic activity (e.g. mining). These are similarconcepts to passive dispersal and active trans-formation, defined by Lewin & Macklin (1987)for mining-contaminated river systems (see CaseStudy 3.2). Bridge’s (2003) dynamic equilibrium–disequilibrium concepts are well-illustrated inFig. 3.9, where a system, previously at equilib-rium, is affected by increase in discharge that

Fig. 3.7 Photograph of typical fluvial overbank sediments, with gravel base and fining upward sequence. Length of tape measure is 1.4 m.

ESC03 9/7/06 3:16 PM Page 83


causes disequilibrium between channel width anddischarge. There is a lag time before the systemreadjusts to an equilibrium state once again.

Other authors have described activation thres-holds for sediment transfer. In these situations,some sort of barrier must be overcome beforesediment transport can occur. Rowan et al. (2000)described two large flood events that causedconsiderable sediment transfer in the river basin of Wyresdale Park Lake in north-west England.They argued that the activation threshold to this transfer was the accumulation of sedimentwithin colluvial footslopes during a dry periodimmediately preceding the flooding. Once thesethreshold-crossing events occur, a river channelresponds, often rapidly and dramatically, and anew set of conditions (changes in gradient, cross-sectional and planform geometry) develops. The remainder of this chapter considers manyexamples of responses to river sedimentation tosuch threshold-crossing events and their under-lying environmental causes.

3.3 PROCESSES AND IMPACTS OF NATURAL

DISTURBANCE EVENTS

Fluvial environments are subject to fluctuationsin rates of sediment transport and accumulationthat are caused by natural disturbance eventsincluding climate change, tectonic uplift, glacio-isostatic rebound, forest fires and volcanic erup-tions. This section presents an overview of theeffects of these processes on sedimentation influvial environments.

3.3.1 Climate change

Climate change can be defined principally asmodifications in precipitation and temperature(and related features such as permafrost cover),and some authors also include modifications to vegetation cover in the definition (Knighton1998). Climate change can occur on a wide rangeof time-scales: from 103 to 105 years (majorglaciations), to 103 to 102 years (interglacial or

Soft rocks

Hard rocks

Erosion andremobilization

Alluvialstorage

Within channelstorage

Colluvialstorage

Loessic input

Deeper lowlandsoils erosion

Infilledpalaeolake

Lakeinfilling

Streamerosion

Colluvial storage

Primaryerosion

Slope processescreep, landsliding,slope, wash, etc.

Episodicerosion

Dry valleyfill

Increasing sediment in storage

103–104103–104

103103 102–103

1–102

102–103

Fig. 3.8 Duration (in years) and location oflong-term sediment storagein a typical humid, cool riverbasin. (After Brown 1997.)

Equilibrium Disequilibrium Equilbrium

Channelwidth

Waterdischarge

Time

Fig. 3.9 Definition of equilibrium anddisequilibrum in stream geometry. Channel widthvaries congruently with water discharge (i.e.equilibrium) until extreme discharge causes bankerosion and channel widening. Subsequently, thechannel is in disequilibrium with water dischargeand requires a certain lag time to regain, bydeposition, equilibrium. (After Bridge 2003.)

ESC03 9/7/06 3:16 PM Page 84


Climatic Optimums such as the Little Ice Age inthe sixteenth to nineteenth centuries), to 101 years(global warming trends in the late 1990s andearly 2000s). These climatic changes are relatedto factors such as Milankovitch cycles (Lewis et al. 2001), changes in meridional circulation(Knox 1995) and global warming, and they play a major role in influencing weathering pro-cesses and, ultimately, channel morphology andchannel and floodplain sedimentation.

Many investigators have demonstrated thatrivers are particularly sensitive to changes in climate, and several have sought to define therelationships between these changes and sedi-mentation in river systems. Early work suggesteda direct relationship between sediment yield andprecipitation (Fig. 3.10a) (Langbein & Schumm1958), in that the highest yields occurred dur-ing peak periods of precipitation. Later work byWalling & Kleo (1979) showed that this rela-tionship was more complex on a global scale(Fig. 3.10b), with a three-peak average sediment-yield–precipitation plot that reflected distinct

global climatic zones. Walling & Kleo (1979)did, however, suggest that this might be too simplistic, and that other factors, such as sea-sonal effects on precipitation and temperature,relief, soil and rock type, and land use, may be responsible for the three peaks on Fig. 3.10b(Hooke 2000). In terms of floodplain deposition,increased discharges have been related directlyto the area of floodplain that is inundated bysuspended-sediment-bearing water (Hamilton1999). The situation is not always as simple asthis: Aalto et al. (2003), for example, showedthat crevasse-splay deposits in the Bolivian Beni and Mamore river basins in Bolivia couldbe grouped temporally, and that the deposi-tion pattern was punctuated in a stop-and-start manner. Aalto et al. (2003) related thesecrevasse-splay deposit groupings to the cold (LaNiña) phases of the ENSO (El Niño–SouthernOscillation) cycle, in that rapidly rising floodsduring these phases destabilize and depositcolossal volumes of Andean sediment. Inman & Jenkins (1999) demonstrated that the wetperiods of alternating, decadal-scale ENSO-induced climate changes were responsible forannual sediment fluxes that were up to 27 timesgreater than those in dry periods.

Flood frequency plays a major role in definingpatterns of river channel erosion and deposi-tion downstream (Rumsby & Macklin 1994).Correlations of alluvial chronologies in the UK,Europe and North America have revealed majordiscontinuities in the Holocene alluvial record,marked by alternating wetter and drier phases,and alternating higher and lower frequency of extreme flood events (Starkel 1983). Thisclustering has been linked mainly to climaticallydriven changes (Macklin & Lewin 1993, 2003;Rumsby & Macklin 1994; Knox 1995). The highflood-frequency clusters resulted in landslides,increased rates of deposition and lateral channelchange and avulsion (Starkel 1983). Because ofthe intimate link between fluvial sedimentationand climate, many investigators have used thefluvial sedimentary record to reconstruct pastclimates (despite the sometimes sparse nature ofthe climate record), particularly in contemporarytimes when global warming is of such concern.

Desertshrub Grassland Forest

Reservoir data

Stream data

0 500 1000 1500

200

400

600

800

Effective precipitation, mm

Sed

imen

t yie

ld, t

km

−2yr

−1S

edim

ent y

ield

, t k

m−2

yr−1

0

Mean annual precipitation, mm

0 800 1600 2400 3200 40000

400

800

1200

(a)

(b)

Fig. 3.10 Relationship of sediment yield to (a) effectiveprecipitation (after Langbein & Schumm 1958; Knighton 1998)and (b) mean annual precipitation (after Walling & Kleo 1979;Knighton 1998), based on measured data.

ESC03 9/7/06 3:16 PM Page 85


In the UK, Macklin & Lewin (2003) identified14 major flooding phases, lying within the timeperiods 400–1070, 1940–3940, 7520–8100 andc. 10,420 cal. yr BP. Although these are mainlylinked to variations in climate, forest clearanceand subsequent agricultural activity (see sec-tion 3.4.1.1) were thought to prime the soils andsediments for erosion and subsequent redis-tribution by the floods, followed by alluviation(Macklin & Lewin 1993, 2003) (Fig. 3.11).

Base-level changes, which are often linked to climatic change (as well as tectonic uplift), are responsible for the accommodation space offluvial sediments. In north-east Japan, low ratesof base-level rise have been related to the develop-ment of large-scale channel fills, while rapidrates have been linked to small-scale channels(Komatsubara 2004). A drop of about 22 m in themean level of the Dead Sea over the past 70 yearsresulted in a constant adjustment of the LowerJordan River (Hassan & Klein 2002). Here, rapiddrops in sea-level, particularly since the 1980s,led to deep incision, changes in channel morpho-logy and terrace development. Stable base-levelsare, in places, related to channel aggradationand only minor lateral migration.

Even though climate change operates over bothlong and short time periods, fluvial responses toit are often rapid within these periods. Goodbred(2003) demonstrated that the Ganges River

system in Bangladesh and India responded to < 104-year climate change in a system-wide, con-temporaneous manner, with punctuated, rapidsediment transfers. Lewis et al. (2001) discussedclimatic shifts within the last interglacial–glacialcycle (125–10 ka), suggesting that these wererapid and that they were recorded in alluvialsediments. They described the changes in theThames, in which major changes in the alluvialarchitecture were related to climatic fluctuationsat c. 70 ka and 13–11 ka (the Devensian Late-glacial climatic warm–cold–warm oscillations).

During Pleistocene cold stages, rivers innorthern Europe and America were similar tocontemporary cold-climate tundra rivers, withoverbank deposition in large, unconfined valleys,rapid vertical aggradation of sandy material onfloodplains and limited lateral channel migra-tion due to well-established tundra vegetation andriver bank cohesion (Kasse 1998). In areas withlittle vegetation cover, large-magnitude springsnow melts would have led to enhanced sedi-ment erosion, transport and deposition. Thicksequences of river gravels in northwest Europe(Antoine 1993; Bridgland 2000) are thought tohave aggraded in such areas, as river environ-ments changed from incisional to depositional(Bridgland & Allen 1996). Cooling stages areoften recorded in the fluvial archive by channelchanges: in the Nochten area, eastern Germany,

Calendar age in kyrs

Drif

t ice

inde

x (%

)

0 2 4 6 8 10 120

5

10

15

British flood episodes

Climatic deteriorations inferred from mire wet shifts in Britain

Upper Mississippi flood episodes, USA

Lake Tutira storm periods, New Zealand

Central European cold-humid phases

Holocene records of North Atlantic drift ice

LIACE CE CE CE CE CE CE CE8 7 6 5 4 3 2 1

Fig. 3.11 Comparison of BritishHolocene flood episodes with proxyclimate records for Britain, centralEurope and the North Atlantic, well-date storm periods for LakeTutira, New Zealand, and floodepisodes in the upper Mississippibasin, USA: LIA, Little Ice Age; CE, central European cold-humidphases (from Haas et al. 1998).(Figure after Macklin & Lewin 2003.)

ESC03 9/7/06 3:16 PM Page 86


a strong cooling event at c. 40 ka led to a changefrom sandy anabranching to braided river con-ditions (Kasse et al. 2003). With deglaciation atthe Pleistocene–Holocene transition came highrunoffs and sediment loads, and considerablechannel change. Thomas (1998) suggested thatthe change from dry to much wetter climaticconditions between c. 1200 and c. 9000 yr BPresulted in high discharges in both mountainous(e.g. Nile, Amazon) and lowland tropical rivers(e.g. Niger) and accompanying landsliding andchanges in slope hydrology and processes. Asvegetation re-established, slopes stabilized anderosion became more confined to bank areas.

During the Holocene, climate change has continued to significantly affect fluvial activity.Alluvial deposits in England are thought to haveformed during the Late Medieval Warm Periodand Late Medieval Deterioration in response tocoincident heavy rain and snowmelt, whereasthose from corresponding periods in Italy arerelated to increased cyclonic activity (Brown1998). Episodic flooding events have been shownto play a major role in present-day sedimenta-tion in the Eel River, California (Morehead et al.2001), and in Scottish rivers, where rates of lateral channel shift and extent of bare gravelsare related to changes in flood frequency since themid-nineteenth century (Werritty & Leys 2001).

3.3.2 Tectonic uplift and glacio-isostatic rebound

Tectonic uplift often causes river incision andshedding of sediment from upland areas to riverbasins. It can result in considerable channelchange, particularly in valley slope, width, dis-charge and, ultimately, denudation and alluvi-ation downstream. The Bengal Basin in India andBangladesh, for example, is rapidly aggradingdue to Himalayan uplift and erosion, and sea-level changes since the last glaciation (Allison et al. 2003). Lave & Avouac (2001) used sus-pended sediment loads in Himalayan rivers to infer that contemporary hillslope erosionover the whole Himalayan range in Nepal wasdriven mainly by fluvial incision linked directlyto uplift. In the Miocene, the gradual tectonicuplift of Africa has been linked to sedimenta-

tion in the Congo River arising from erosion of the edge of the uplifted Angolan margin(Lavier et al. 2001). In the Tejo river, Portugal,successive regional uplift events during increasedintra-plate compression are thought to causedramatic fluvial incision (Cunha et al. 2005). Inthis area, river terraces, previously thought to be related mostly to glacial–interglacial cyclessuperimposed upon uplift, were reinterpreted asbeing caused mainly by these tectonic phases,punctuated by relatively short periods of lateralerosion and some aggradation, with climaticand base-level change factors only acting as aconditioning process.

Tectonic uplift can result in a short-term orgradual change in longitudinal profile, particu-larly when the uplift is domal (Knighton 1998).In the eastern Carpathians, Radoane et al. (2003)demonstrated that tectonic uplift at over 6 mmper year, rather than age, was responsible for theshape of the longitudinal profile and the types ofchannel deposits present. Laboratory experi-ments carried out by Ouchi (1985) have shownmeandering rivers respond to slow uplift byincreasing sinuosity (in the case of a steepeningvalley slope), or by straightening or anastomos-ing (in the case of a flattening slope).

In north-west Europe, uplift as a result of glacio-isostatic rebound is thought to play a role incontrolling Early Pleistocene sedimentation inrivers (Bridgland 2000; Westaway et al. 2002).The basis of this theory is that, to maintain iso-static equilibrium, a particular area will undergouplift when adjoining areas undergo repeatedand cyclical glacial surface loadings (Westawayet al. 2002). This theory has been used to explainthe occurrence of thick (c. 100 m) sequences ofgravel terraces, aggraded during glacial cycles.Although climate is the major factor responsiblefor these terraces, in that large-magnitude sea-sonal flows and reduced vegetation cover causeincreased erosion and large-scale deposition(Bridgland & Allen 1996), the shedding of thissediment requires uplift during interglacial marinehighstands (Bridgland 2000; Westaway et al.2002). There may also be an element of mantlebulge or upwarp involved in response to crustalunloading as a result of terrain erosion.

ESC03 9/7/06 3:16 PM Page 87


Burger et al. (2001) also made a link betweenriver morphology and glaciation-related sea-levelchanges. They suggested that the dendritic channelpattern (and related fluvial incision) in the EelRiver Basin, California formed during glacio-eustatic lowstands and periods of continental shelfexposure, but also pointed out that local tectonicuplift may have also played a role. In fact, manyother studies (e.g. Veldkamp & Van Dijke 1998)have demonstrated the complex links betweentectonic, climatic and sea-level changes and theircombined effects on river morphology and sedi-mentation. Although Harvey et al. (2003) showedthat sedimentary sequences and the morpho-logical evolution of late Quaternary alluvial fansin the Tabernas basin of south-east Spain weremainly controlled by climatic change, the de-positional environment of toes of the fans wascreated in response to tectonic uplift. Bridg-land (2000) and Bridgland & Maddy (2002)suggested that sedimentation–incision cycles ofthe Middle–Late Pleistocene that correspond to the low-frequency, high-amplitude 100 kyreccentricity-driven climate cycle resulted fromthe overprinting of long-term fluvial responsesto tectonic uplift driven by climate changes insediment supply.

3.3.3 Forest fires

Forest fires result in the removal of vegetation,reduction of ground cover and alteration of soil properties (e.g. decreased infiltration rates, formation of water-repellent layers by vaporiza-tion of forest organic compounds). All of theseprocesses can cause significant increases in run-off and fine sediment production in the fire-affected basins (Morris & Moses 1987; Inbar et al. 1998). In Australia, for example, hydrogeo-morphological changes resulted in widespreaderosion and alluvial deposition of remobilizedtopsoil in river systems following the Christmas2001 bushfires near Sydney (Shakesby et al.2003). Many other studies have demonstratedsimilar widespread post-fire sediment deliveryto rivers (e.g. Inbar et al. 1998) (Fig. 3.12).

The fluvial response to forest fire is complex(Schumm & Lichty 1965), and depends on the

extent and severity of the fire, the rate of recoveryof vegetation and the fluvial geomorphology. It has been suggested that the most importantfactor is the percentage of burned area, becausethis causes a range of fluvial adjustments through-out the whole basin (Legleiter et al. 2003).Following a fire, the width, depth, planimetricgeometry and longitudinal profile of the riverwill be modified to accommodate the elevatedsediment loads arising from the increased flowmagnitude and frequency. Channels are reportedto respond to fire-related change by aggrada-tion, active braiding, enlargement and lateralmigration, entrenchment and narrowing (Laird& Harvey 1986; Legleiter et al. 2003). The timeneeded for basins to return to pre-fire surfaceerosion and fluvial sediment delivery rates hasbeen reported to be relatively short, in the orderof 2 to 10 years (Morris & Moses 1987; Legleiteret al. 2003). This normally follows exhaustionof available fine-grained sediment, vegetativerecovery and the development of coarse surfacelags (Morris & Moses 1987).

It is important to note that, although forestfires are often ‘natural’ events, they can equallybe classified as ‘anthropogenic disturbances’,because many forest fires are started deliberatelyby humans. This is not only a contemporaryphenomenon; late Mesolithic to early Neolithiccharcoal finds in the North Yorkshire Moors,UK, suggest that fires were started deliberatelyto control vegetation or drive out game forhunting, or both (Simmons & Innes 1996).

3.3.4 Volcanic eruptions

Explosive volcanic eruptions can dramaticallyaffect sedimentary processes in rivers, both dur-ing the eruption itself and for years or decadesafter the eruption has ceased. The sedimentloads of rivers draining volcanoes are among the highest documented, and have been comparedwith those of rivers in arid climates impacted by flash floods (Hayes et al. 2002). The effects of high rainfall in volcanic areas (particularlytropical volcanic belts) compound the sediment-carrying capacity of associated rivers comparedwith arid rivers. During eruptions, river valleys

ESC03 9/7/06 3:16 PM Page 88


can be rapidly aggraded with great thicknessesof volcanic deposits (ash, pumice, bombs). The1991 Mount Pinatubo eruption in the Philippines,for example, resulted in infilling of valleys with200 m of volcanic debris (Scott et al. 1996). Themost devastating effects on rivers, however, arisefrom lahars (Cronin et al. 1997; Hayes et al.2002). Lahars are hyperconcentrated to concen-trated volcanic debris flows, which move at highspeed downslope and can reach considerabledistances downstream. They are caused by pyro-clastic (hot) flows combined with snow melt(Fig. 3.13), or heavy syn- or post-eruptive rain-fall (Neall 1976). Lahar formation is favouredby high slope gradients, removal of vegetationdue to eruptive activity, an abundance of uncon-solidated volcanic ejecta and, in many cases,pre-existing channel and channel bed morpho-logy (Hayes et al. 2002). Channel and floodplain

sediments in volcanically affected rivers are incisedand reworked relatively soon (often within amatter of years) after deposition (e.g. Janda et al. 1996). This can result in the deposition of alluvial fans and in destruction of propertywhere population centres lie downstream. Forinformation on volcanically triggered sedimen-tation events see Chapter 2.

3.4 PROCESSES AND IMPACTS OF ANTHROPOGENIC

ACTIVITIES

3.4.1 Anthropogenic impacts on rates and stylesof sedimentation

Human exploitation of natural resources fromprehistoric times through to the present day hashad a very significant impact on the hydrology,

Burned drainage area (%)

Cro

ss-s

ectio

nal s

trea

m p

ower

(W

/m)

0 10 20 30 40 50 60 70 80

10

1

00

10

00

100

00

1989/90 1990/91 1991/92 1992/93* 1 2 3 4 5 6 Years

103

102

101

100

10−1

10−2

Measured yield in experimental plots

Assumed trend

Baseflow sediment yield (control plot)

Fire-induced 'accelerated' sediment yield

Sed

imen

t yie

ld g

/m2

(a)

(b)

Fig. 3.12 Responses of rivers to forest fires.(a) Relationship between cross-sectionalstream power and percentage of burneddrainage area. The solid line represents theregression line and the dashed lines representthe minimum and maximum cross-sectionalstream power values. Note the importantincrease in minimum stream power for moreextensively burned basins. (After Legleiter et al.2003, fig. 6.) (b) Fire-induced sediment yieldagainst ‘baseflow’ yield in the Mount Carmelarea, Israel. The 1992 increase in sediment yield is due to exceptionally heavy rains. (After Inbar et al. 1998, fig. 17.)

ESC03 9/7/06 3:16 PM Page 89


sediment regime and ecology of many of theworld’s rivers. Deforestation, agriculture, land-use change, river regulation and mining are someof the anthropogenic activities that affect soil

erosion rates and sediment conveyance in riversystems. Increases in flooding and flood peaks,river sediment loads and rates of valley-flooralluviation following human disturbance, or dis-turbance of vegetation, are now well documented(Robinson & Lambrick 1984). This section dis-cusses several types of anthropogenic impactson sedimentary processes in river systems.

3.4.1.1 Agriculture, deforestation and afforestation

Human settlement and cultivation have had adrastic impact on river sediment budgets, largelythrough deforestation and land clearance. Theseactivities result in increased soil erosion, soilcreep and landslide events, which in turn resultin increased sediment input to, and sedimenta-tion rates in, rivers (Fig. 3.14) (Knox 2001).Furthermore, the sediment composition is oftenaltered, because progressively less weatheredand less organic-rich terrigenous materials arestripped and added to the suspended load dur-ing clearance.

Fig. 3.13 Destruction of Armero, Columbia, by a lahar,triggered by an eruption and melting snow, which emanatedfrom the Nevado del Ruiz volcano. More than 23,000 peopledied during this event. (Image and information courtesy of R.J. Janda, USGS/Cascades Volcano Observatory.)

0 10 20 30 40 50 60

100

98

96

94

92

90

90

80

70

60

50

40

30

20

10

01820 1840 1860 1880 1900 1920 1940 1960 1980 2000

Year

Distance (m)

Ele

vatio

n (m

)U

nits

(se

e le

gend

)

Discharge (cm)Thalweg shear stress (N m−2)Channel capacity (m2)

Historical channel evolution

1820 1890 1925 1940 1990

Fig. 3.14 Historical changes inchannel morphology and hydraulicson the Shullsburg Branch tributary,Galena River system, USA.Accelerated overbank sedimentationfrom the beginning of Euro-Americansettlement in the 1820s increasedprogressively, resulting in increases ofbank heights and facilitation of deeperflows with high shear stresses that led to bank erosion and channelexplansion. After about 1950,overbank sedimentation decreasedowing to smaller and less frequentfloods as a result of improved upland land conservation practices.(After Knox 2001, fig. 9.)

ESC03 9/7/06 3:16 PM Page 90


Numerous studies have demonstrated increas-ing sediment loads to rivers since human settle-ment. Mei-e & Xianmo (1994) described anincrease in soil erosion in the Yellow River, China,since serious cultivation began c. 5000 yr BP.Following another phase of clearance c. 1000 yrBP, Saito et al. (2001) calculated that the sedi-ment load of the Yellow River at Sanmenxiaincreased by four to seven times more than itspre-1000 yr BP level of 2.7–5.0 × 108 t yr−1. Inthe USA, sediment loads of rivers draining to theAtlantic have been estimated to have increasedfour to five times since European settlementand, in particular, high sedimentation rates inChesapeake Bay and the Appalachian Piedmonthave been linked to agricultural clearance in the Appalachian mountains in the nineteenthcentury (Kearney & Stevenson 1991). Brown(1997) discusses a plethora of examples ofanthropogenically induced alluviation in the UKand rest of Europe. Among these is the UpperThames, where increased alluviation is ascribedby Robinson & Lambrick (1984) to coupledbuilding of large villas and hillslope agricul-ture during the Roman Occupation of Britain.Widespread mid- to late-medieval alluviation in Britain and a considerable part of northernEurope has also been attributed to populationincreases and increased ploughing during thisperiod (Brown 1997).

Deforestation and related sediment inputs to rivers also cause significant channel change.Coltori (1997) attributed aggradation and forma-tion of a braidplan system in the Marche basinto deforestation of the Periadriatic Basin andsubsequent severe soil erosion. Post-deforestationlandslides have made significant impacts on the sedimentary sequences of several deposi-tional basins in New Zealand, and in Europedeforestation has also resulted in landslides thatwere triggered by high-magnitude–low-frequencyclimatic events, causing gully and channel ero-sion (Glade 2003). Changes in river channelwidth, sinuosity and position in basins at thefoothills of the Venezuelan Andes were greatestat the most intensely deforested sites, althoughvalley shape and other constraints on channelswere important (Karwan et al. 2001).

The effects of agriculturally related landclearance are not always felt immediately. In the Chagrin River, Ohio, USA, high sedimentyields matched river basin deforestation duringthe 1840s to 1850s, but rates declined by over50% from the 1980s to the 1910s, despite peaksin land clearance (Evans et al. 2000). Betweenthe 1910s and 1960s, declines in farming cor-responded to increases in sediment yields. Thesetrends were attributed to the timing of peakflows, and the resultant intrabasinal storage anddelays in sediment conveyance from upstreamto downstream. Walling (1999) described thisdelay as the ‘buffering capacity’ of a basin, andrelated it to the sediment delivery ratio, whereasKnox (2001) described the delay as a ‘long-termlag response’.

Soil erosion is also promoted through agri-cultural practice, and currently poses a threat to the sustainability of agricultural productionin many areas of the world. Remobilization oferoded soil causes siltation of irrigation net-works and sedimentation of reservoirs that areused for water storage. It has been estimated that75 billion tonnes of soil are eroded annually,and that most of this comes from agriculturalareas (Myers 1993). Speth (1994) suggested that approximately 80% of the world’s agricul-tural areas are undergoing ‘moderate to severeerosion’, and a further 10% from ‘slight to moderate erosion’. Most of this eroded soil willchannel through river basins. Rates of erosion,however, are highest in Asia, Africa and SouthAmerica (1000–2000 t km−2 yr−1) and lowest in Europe and the USA (c. 0–600 t km−2 yr−1;Fournier 1960).

Agricultural and related practices also con-tribute to contamination of fluvial sediments, par-ticularly with respect to fluxes of the nutrientsC, N and P. Owens et al. (2001), for example,suggested that elevated particulate P concentra-tions in the upstream River Aire, UK, were pro-bably related to diffuse sources of P connectedwith agricultural land-use. In the Amazon, pro-gressive deforestation over the past few decadeshas disrupted the fluxes of humus and soil fromthe drainage basin, because organic-rich mater-ial that is normally retained in the soil is being

ESC03 9/7/06 3:16 PM Page 91


progressive weathered and mobilized (Farella et al. 2001). Recently deposited alluvium is thusmore humified and N-rich than older material.

Many authors (e.g. Macklin et al. 1992;Macklin & Lewin 1993; Brown 1997; Coulthardet al. 2000) have stressed that anthropogenicactivity and climate change act in tandem toaffect sedimentation in river basins. In a numer-ical (computer) modelling study, Coulthard et al.(2000) showed that, although deforestationalone increased sediment discharges by 80% inan upland basin in northern England, a changein both climate and vegetation cover resulted ina 1300% increase, confirming field-based studiesin the area. River basins are therefore more sensitive to climate change when the soils aredestabilized by human activities such as defore-station and agricultural practices (Coulthard et al. 2000). These activities expose relativelyfresh, unweathered material, making it moresusceptible to, and possibly inflating the ratesof, chemical and physical weathering.

Afforestation, particularly in upland portionsof basins, can also have significant effects onsedimentation in rivers, owing to soil distur-bance when planting. Among the effects docu-mented are (i) elevated rates of bank erosion(Painter et al. 1974), increased suspended sediment yields, increased bed load yields, par-ticularly upstream (Newson 1980), and develop-ment of a flashy hydrological regime (Leeks &Roberts 1987). Mount et al. (2005) constructedsediment budgets to evaluate the impacts ofupland afforestation between 1948 and 1978 inthe Afon Trannon, mid-Wales. These budgetsdemonstrated that upland basin bed load yieldsof between 2 and 3 t km−2 yr−1 were equivalentto localized gravel inputs from bank erosion,and thus were probably not responsible forincreased sedimentation downstream. Rather,the nature of the bank material, and increases inflood magnitude and frequency since 1988, pos-sibly as a result of climate change, were deemedto be responsible. This study again demon-strates the interplay between different naturaland anthropogenic processes in influencing riversedimentation, and the need for interdisciplin-ary study to understand these processes.

3.4.1.2 Mining

Present-day and historic mining activities havereleased large quantities of metal-bearing mineralparticles into fluvial environments (Fig. 3.15a;Lewin et al. 1977, 1983; Lewin & Macklin1987; Macklin et al. 1994; Hudson-Edwards et al. 1996, 1999b, 2001; Miller 1997). Mining-related sources of metal-bearing sediments torivers include discharge of mine or processingwaste, tailings dam failures, erosion of tailingsand waste rock piles, remobilization of mining-contaminated alluvium, and mine drainage.Direct discharge of mine tailings, effluent andwaste rock is one of the most common andsignificant sources of metal-bearing particles to river systems. This has been an importantprocess in the past (e.g. north-east England;Hudson-Edwards et al. 1996), but is still ongoing today (e.g. Río Pilcomayo, Bolivia;Hudson-Edwards et al. 2001; Fig. 3.15b). Oreprocessing activities (e.g. amalgam treatment,cyanide leaching) also contaminate rivers withmetals and metalloids such as Ag, Bi, Cd, Cu, Hg, Pb, Sb and Zn which, over time, can accumulate in significant amounts (Fig.3.16).

Failures of tailings dams can release very largequantities of metal-bearing particles to river systems. The discharge of these tailings oftengreatly exceeds the sediment-transport capacityof a river and results in considerable channeland floodplain aggradation (James 1989). Phys-ical remobilization of abandoned tailings orwaste piles, and of channel beds and mining-contaminated floodplain alluvium (formed dur-ing historic mining activity), also provides largeamounts of metal contaminants to rivers (Macklinet al. 1992; Miller 1997; Miller et al. 1998).Merrington & Alloway (1994), for example,showed that the wash load being transferredfrom abandoned tailings heaps in two mines inCornwall and Wales, UK, was very high for Cu(38 kg yr−1) and Pb (74 kg yr−1). Erosion occursduring lateral channel migration and exposureof bank sediment (Miller et al. 1998), or channelbed, tailings, waste pile or cut-bank incision(Macklin & Lewin 1989).

ESC03 9/7/06 3:16 PM Page 92


Fig. 3.15 Mine waste in the Río Pilcomayo, Bolivia. (a) Braided reach, in which the entire valley is filled with mining-related sediment.The river is grey-coloured and contains sulphide mine tailings, and in places on the floodplain the sulphides have oxidized to formmetal-bearing iron oxyhydroxides. Approximate width of photograph is 500 m. (b) Discharge of sulphide mine tailings directly into the river.

(a)

(b)

ESC03 9/7/06 3:16 PM Page 93


Mine drainage from both metalliferous andcoal mining areas is a common source of metalcontamination to river systems. The leaching of metals from exposed and buried waste is controlled by factors such as mineral dissolu-tion, sulphide mineral oxidation and secondary mineral precipitation. These secondary mineralsoften precipitate or flocculate as discrete particlesor as coatings on other grains; these particles co-precipitate or sorb the dissolved metals andare subsequently deposited or transported down-

stream. Mining operations themselves can alsoalter the natural fluvial geomorphology duringthe construction of artificial drainage networksand other structures.

3.4.1.3 River regulation and channelization

River systems are regulated for many reasons,including more efficient navigation, flow, bankerosion, drainage and flood control, stabiliza-tion of channels, provision of surface water

0.0

0.4

0.8

1.2

1.6

2.0

2.4

<10 km2 10−50 km2 50−100 km2 >100 km2

Drainage area (km2)

Sed

imen

tatio

n ra

te (

cm y

r−1)

Historical overbank sedimentation rate

Pre-miningMiningPost-mining

0.01 0.1 1

Pre

hist

oric

His

toric

Au Hg OC

Dep

th (

cm)

0

100

200

300

400

Au & Hg (ppm); OC (%)

(a)

(b)

Fig. 3.16 Mining-related accumulation of metals infloodplains. (a) Vertical distribution of gold (Au),mercury (Hg) and organic carbon (OC) in floodplainalluvium, north Georgia. (After Leigh 1997.) (b) Historical overbank sedimentation rate of pre-, syn- and post-mining alluvium in the Blue RiverWatershed, Wisconsin. Note the spatial and temporallag in redistribution of syn-mining alluvium in thedownstream (> 100 km2) portion of the basin. (After Lecce & Pavlowsky 2001.)

ESC03 9/7/06 3:16 PM Page 94


reservoirs for drinking, irrigation or other pur-poses, hydroelectric power, removal of naturalvegetation and waste disposal (Newson 1992).Typically, regulation is carried out by dam andreservoir building, and channelization, whichcan involve resectioning, sediment extraction,the use of flood walls, culverts, embankments orlevees, and straightening.

Sedimentation in regulated rivers can be con-siderably different from that in natural rivers. In dammed rivers, reservoirs act as sedimenttraps, significantly affecting rivers with mediumto large natural sediment loads. Downstream ofdams, many rivers have been reported to supplyless sediment after damming for irrigation, floodcontrol and power generation. This reductionhas resulted in catastrophic changes in down-stream river reaches, and deltas and coastlines,which are often eroded as a result of increasedstream power (Trenhaile 1997). In the Missouri

and Mississippi rivers, for example, the reductionof annual sediment yield from 817 million tonnesto 204 million tonnes in the twentieth century isthought to result from large-scale dam projects.This loss in sediment yield is in turn linked torapid coastal degradation and recession in down-stream Louisiana (Keown et al. 1986). Walling& Feng (2003) carried out an extensive study oflong-term records of annual sediment load andrunoff for 145 major rivers, and demonstratedthat although c. 50% of the rivers experiencedstatistically significant upward or downwardtrends, most showed the latter. This was largelyattributed to dam construction. Impacts of theAswan High Dam on the River Nile are also dis-cussed in Chapter 7, and the Hoover Dam inColorado in Chapter 1.

Many studies have demonstrated effects onriver sedimentation as a result of river regulation(e.g. Case Study 3.1). Arnaud-Fassetta (2003)

Case Study 3.1 Past, present and future impact of the Three Gorges Dam on the Yangtze River

The Yangtze River (Changjiang) is the longest river in Asia (> 6300 km), the third longest in the world and has the ninth largest catchment basin (1.8 × 106 km2) in the world. Its waterdischarge is the largest in the western Pacific Ocean, and its sediment load (annual discharge of480 × 106 t yr−1 before the 1990s) is the third largest in the world behind the Amazon and theCongo. Shanghai, the most important industrial and economic city of China, is located near theriver mouth. The 39.3 × 109 m3 Three Rivers Gorge Dam project, scheduled to be completedbetween 2009 and 2012, will be the largest dam in the world, with estimated power productionequalling approximately 12 nuclear power plants. The dam is being built for flood manage-ment, inland navigation and hydropower.

Yang et al. (2002) reported that decadal river sediment discharge and suspended sedimentconcentrations (SSC) had reduced by 34% and 38%, respectively, between the 1960s and 1990s(Case Fig. 3.1a). This was in contrast to the period between the 1950s and 1960s, when annualsediment discharge and SSC increased by 10 and 12%, respectively. Yang et al. (2002) attributedthese increases to deforestation, and the 1960s to 1990s decreases to a combination of dam andreservoir construction (Case Figure 3.1b). The authors used these data to predict that both riversediment discharge and SSC will probably decrease to 40% of original levels in 50 years and to50% in 100 years, largely as a result of the dam and reservoirs. Critics have suggested that thedamming will also lead to the build-up of sediment-borne contaminants, damage of turbines bysediment, increases in deforestation and soil erosion as former inhabitants relocate and erosionin the delta downstream of the dam. Wang et al. (2005) have demonstrated that, after the com-pletion of the dam, anastomosing channels downstream of the site will cease to carry significantamounts of water and sediment, and thus disrupt the balance between fluvial discharge and

ESC03 9/7/06 3:16 PM Page 95


showed that engineering of the Rhône River inFrance, carried out to control overbank flood-ing and stabilize the river planform, resulted inchannel narrowing and entrenchment, and in-creased hazard of low-frequency, high-magnitudeflood events (Fig. 3.17a). Surian & Rinaldi(2003) carried out a comprehensive review ofthe response of Italian rivers to sediment extrac-tion, dams and channelization, and highlighted

channel incision, generally 3–4 m but up to 10 m, and channel narrowing of up to 50%, asthe major effects (Fig. 3.17b). They did, how-ever, point out that these effects were most pronounced immediately after the engineeringworks were carried out, and that the effects slowedand became asymptotic. Gaeuman et al. (2005)demonstrated that the creation of large-scalediversions and increased water withdrawal for

basin subsidence due to the Himalayan orogeny. By contrast, others suggest that a decline in sediment supply downstream of the dam will be beneficial in that flood levels will be less.Furthermore, the Danjiangkou Dam, built on a tributary of the Yangtze, is not believed to havecaused downstream incision, and concerns that coastal erosion will occur are countered by suggestions that mud flats north of the Yangtze will be supplied by tributaries downstream ofthe Three Rivers Gorges. The full effects of the Dam Project, and answers to these questionswill not, however, be available for many decades.

Relevant reading

Fuggle, R., Smith, W.T., Hydrsult Canada Inc. & Agrodev Canada Inc. (2000) Large Dams in Water andEnergy Resource Development in the People’s Republic of China (PRC). Country review paper prepared asan input to the World Commission on Dams, Cape Town. http://www.dams.org (accessed 25 June 2004).

Milliman, J.D. & Meade, R.H. (1983) World-wide delivery of river sediment to the oceans. Journal of Geology91, 1–21.

Wang, S., Chen, Z. & Smith, D.G. (2005) Anastomosing river system along the subsiding middle Yangtze Riverbasin, southern China. Catena 60, 147–63.

Wang, Y. & Zhu, D.K. (1994) Coastal Geomorphology. Higher Education Press, Beijing, 244 pp. (In Chinese.)Yang, S., Zhao, Q. & Belkin, I.M. (2002) Temporal variation in the sediment load of the Yangtze river and the

influences of human activities. Journal of Hydrology 263, 56–71.

Calendar year

Ann

ual s

edim

ent

disc

harg

e (1

,000

,000

t)

1945 1955 1965 1975 1985 1995 2005

800

700

600

500

400

300

200

100

01967 1970 1973 1976 1979 1982 1985

0

200

400

600

800

1000

1200

Calendar Year

Am

ount

of

Sed

imen

t(1

,000

,000

t)

sediment fromthe riversediment dammedin the reservoir

(a) (b)

Case Fig. 3.1 (a) History of annual sediment discharge at Datong Station, suggesting a decreasing trend since the 1960s. (b) Accumulative river sediment discharge and amount of sediment dammed in Danjiang Reservoir, middle reach of the Hanjiang River, one of the major tributaries of Yichange and Datong.

ESC03 9/7/06 3:16 PM Page 96


irrigation during the early part of the twentiethcentury in the Duchesne River basin, Utah, causedlocal gully erosion and subsequent downstreamdelivery of fine sediment, which in turn led to

filling of secondary channels, the narrowing of themain channel and transformation of a narrowsand-bed channel into a braided gravel-bed reach(Fig. 3.17c).

A. Petit Rhône River

B. Grand Rhône River

Distance from the sea (km)

Distance from the sea (km)

1907

1999

1895

1995

40 45 50 55

5 10 15 20 25 30 35 40

700

600

500

400

300

200

100

0

350

300

250

200

150

100

50

0

W =

Cha

nnel

wid

th (

m)

W =

Cha

nnel

wid

th (

m)

950 1000 1050 1100 1150 1200 1250

21

17

13

9

5

Progressive distance from datum (m)

Ele

vatio

n (m

a.s

.l.)

1936

1954

1978

(a)

(b)

Sand-bed reaches

Gravel-bed reaches

Gravel-bed reaches

Sand-bed reaches

Q+, Qs−Q−, Qs+

avulsion

large flood

sinuosity−, slope+

QG+ QG

+

(c)

Fig. 3.17 (a) Changes in channel width in the Petit Rhône River (A) and the Grand Rhône River (B) between 1895 and 1999. Bothgraphs highlight the channel narrowing as a result of channelization at the end of the nineteenth century. (After Arnauld-Fassetta 2003.)(b) Channel incision along the Arno River in the Lower Valdarno–Pisa Plain, Italy, showing typical change in cross-section, with limitedbed lowering from 1936 to 1954, and intense incision from 1954 to 1978. (Based on fig. 2 of Surian & Rinaldi 2003.) (c) Summary ofadjustments in sand-bed and gravel-bed reaches of the lower Duchesne River, Utah, USA, owing to the creation of large-scale diversionsand increased water withdrawal for irrigation. Bold arrows indicate the primary adjustment to changes in discharge (Q) and finesediment supply (Qs), and thin arrows indicate second-order adjustments. QG

+ indicates an increase in gravel-sized sediment derivedfrom lateral erosion. (After Gaeuman et al. 2005, fig. 15.)

ESC03 9/7/06 3:16 PM Page 97


3.4.1.4 Urbanization

Urban expansion can have a major effect onchannel morphology and sediment supply anddeposition. Activities such as building on floodplains, increased ground-water extraction,channelization, flood embankment and urban-drainage-system construction, sewage disposaland (potentially contaminated) road runoff areamong those that have major effects. These effectsinclude floods and slope failure, particularly intropical areas (Gupta & Ahmad 1999). Oftenthe maintenance of river channels and floodalleviation systems is extremely costly.

Sediment discharge is often highest during the initial stages of urban development, whenstripping of soil and vegetation to make way for built structures is at a peak. After urbaniza-tion is complete, there is a period of adjustmentas rivers respond to decreased sediment supply.The combination of the removal of soil and vegetation, creation of impermeable surfaces andinstallation of drainage networks in urban areascauses increased runoff during storms, with rela-tively short lag times and high peak discharges(Knighton 1998). More detailed information onurbanized rivers can be found in Chapter 6.

3.4.2 Sediments as sinks for contaminants:transport, deposition and remobilization

3.4.2.1 Sediment contaminants

As discussed in section 3.2.1.1, rivers are subjectto contamination by a large number of metal,nutrient, organic and radionuclide elements andcompounds. A large proportion of the contam-inant load in fluvial systems is transported by particulate matter. Trefry & Presley (1976) andGibbs (1977) suggested that up to 90% of themetal load is transported by sediments in manyrivers, but warned that this varied from metal to metal. This was borne out by Goldstone et al.(1990), who showed that the majority of Znreleased from effluent in Norwich, England,existed principally as dissolved species (95%),whereas Pb was partitioned equally between dis-solved and particulate phases. The partitioning

also appears to apply to organic contaminants, asshown by Lau et al. (1989), who estimated thatchlorinated organic contaminants were almostequally distributed between the soluble and par-ticulate phases in the Detroit River, USA. The‘dissolved’ portion encompasses contaminantsthat are truly dissolved, colloids or sulphide‘clusters’. The latter were described by Rozan et al. (2000), who showed that Fe, Cu and Znwere bound to multinuclear sulphide clusters in oxic rivers in Connecticut and Maryland,USA, which received substantial inputs of sewageeffluent. The ‘particulate’ portion of the con-taminant load comprises contaminant-rich grains(e.g. metal sulphide grains from tailings effluent)or contaminant-bearing Fe and Mn oxide coat-ings on other particles.

In rivers subject to acid-mine or acid-rockdrainage, enormous quantities of metals can betransported downstream in the solute phase(Filipek et al. 1987). In general, however, thissolute phase is considerably reduced downstreamof the source as pH rises to neutral (usually as aresult of tributary or ground-water inputs), andthe dissolved metals are precipitated or adsorbedonto sediments (Davis & Leckie 1978). In rarecases, such as the Río Tinto in south-west Spain,acid pH is maintained along the whole course ofthe river and significant quantities of dissolvedmetals are transported to estuaries where theyprecipitate or are sorbed to other particulates(Hudson-Edwards et al. 1999b). In rivers ofneutral or higher pH, metals and other sub-stances are largely transported downstream inthe particulate load, as discussed above (Gibbs1973; Benjamin & Leckie 1982; Horowitz &Elrick 1987).

The partitioning of contaminants between the dissolved and particulate load in fluvial sys-tems depends on both physical and chemicalfactors. The chemical factors include variationsin amounts of suspended and deposited sedi-ments (Gibbs 1977; Kratzer 1999), adsorptiononto fine-grained material, co-precipitation withor sorption on hydrous Fe–Mn oxyhydroxidesand carbonates, association with organic matter(Karickhoff 1981), incorporation in crystal latticesof minerals, acidification, salinity, complexing

ESC03 9/7/06 3:16 PM Page 98


agents, biomethylation and, probably most signi-ficantly, pH and redox processes (Hem 1972;Gibbs 1977; Salomons & Förstner 1984).

3.4.2.2 Sediment-borne contaminant transport

A large number of physical factors influenc-ing contaminant transport and the partitioningbetween the dissolved and particulate load havealso been identified. These include sediment texture, site geomorphology (Eyre & McConchie1993; Macklin 1996), streamflow, suspendedsediment concentration (Yeats & Bewers 1982;Kratzer 1999), climate and season (Boyden et al.1979; Macklin 1996), water depth, flow dyn-amics (Rust & Waslenchuk 1974; Kratzer 1999),diffusion across the sediment–water interface(McKnight & Bencala 1989) and sediment grainsize (Moore et al. 1987). Gibbs (1977) andHorowitz & Elrick (1987) suggested that grainsize is possibly the most significant factor controlling the concentration and retention ofcontaminants in both suspended and bottomsediment. Metals, in particular, have been shownto be enriched in the fine silt and clay fractionsof sediments (Salomons & De Groot 1978), as aresult of their large surface area, organic and claycontents, surface charge and cation exchangecapacity (Horowitz & Elrick 1987).

Once sediment-bound contaminants enterriver systems, they are dispersed by fluvial pro-cesses and transported downstream. Many ofthe significant advances in the understanding of the dispersal, storage and remobilization ofsediment-bound contaminants in river systemshave been made through studies of metal mining-affected rivers (see reviews by Lewin &Macklin 1987; Macklin 1996; Miller 1997). Theprocesses described in these papers also apply to rivers contaminated by other elements andcompounds (e.g. lipophilic organic compounds,Rostad et al. 1999; radionuclides, Cochran et al.2000; nutrients, Owens et al. 2001). Work onmining-contaminated rivers has demonstratedthat sediment contaminant metal concentrationstend to decrease downstream from pollutionpoint sources in a systematic way that can beapproximated using negative linear, exponential

or power functions (Wolfenden & Lewin 1977;Lewin & Macklin 1987). Deviations from thesemodels are due to floodplain storage or inputs of contaminants from diffuse or other pointsources (Axtmann & Luoma 1991; Macklin1996). The downstream decreases have beenattributed to:1 dilution of contaminated sediment by un-contaminated sediment derived from tributaries,channels and tributaries upstream of contamin-ant point sources and erosion of channel banks(Macklin 1996; Hudson-Edwards et al. 2001);2 hydraulic sorting of channel-bed sediment onthe basis of density, size or shape, which selec-tively enriches sediment near the mining sourcein contaminant metals (Langedal 1997a; Leigh1997);3 abrasion of contaminated sediment grains(Langedal 1997b);4 storage of contaminated particles in channeland floodplain deposits (Macklin et al. 1992);5 chemical sorption or dissolution of contam-inants and/or contaminant uptake by biota(Lewin & Macklin 1987; Hudson-Edwards et al. 1996).Geomorphological and sedimentary processesare influenced by, and control, the dispersion andtransport of sediment-bound contaminants, asillustrated by the definition of two end-memberresponses of rivers to inputs of mine waste (Lewin& Macklin 1987): active transformation andpassive dispersal (see Case Study 3.2).

3.4.2.3 Deposition of sediment-borne contaminants

Within river systems, channels, floodplains,riparian wetlands and reservoirs are the ultimatesinks for metal, nutrient and radionuclide deposi-tion and storage (Macklin et al. 1992; Chesnokovet al. 2000). Many studies have demonstratedthat enormous quantities of contaminants (up toseveral millions of tonnes) can be stored in theseenvironments. Hudson-Edwards et al. (1999a),for example, established that approximately 6.2× 108 t of Pb and 6.4 × 108 t of Zn were storedin 2000-year-old floodplains of the YorkshireOuse basin, UK, which had been affected byindustrial activity and Pb–Zn mining.

ESC03 9/7/06 3:16 PM Page 99


Case Study 3.2 Fluvial responses of rivers to inputs of mine wastes: active transformation andpassive dispersal

The response of rivers to inputs of both contaminated and uncontaminated particles can be viewedas a continuum, with active transformation at one end, and passive dispersal at the other. In activetransformation, the massive input of large quantities of contaminated sediment engulfs the fluvialsystem, causing changes in the types, rates and/or magnitudes of fluvial erosional and depositionalprocesses. This in turn causes the rivers to undergo significant transformations in channel form,which influences the deposition and storage of the contaminated sediment. The classic study ofGilbert (1917) demonstrated that river systems in the Sierra Nevada area, California, respondedto enormous inputs of metal mining waste by aggrading their channel beds by 3–5 m during a10–20 year period after hydraulic gold mining ceased in 1884 (Case Fig. 3.2a(i)). After all of themining-contaminated sediment had been exhausted, the channel beds were slowly reworked anddegraded (over tens to hundreds of years), eventually returning to their pre-mining elevations(Case Fig. 3.2a(i)). Gilbert (1917) modelled the downstream aggradation-degradation processas a simple, symmetrical sediment debris wave. Graves & Eliab (1977) extended the work ofGilbert (1917) and showed that the wave in fact was skewed (Case Fig. 3.2a(ii)); this reflectedthe temporary floodplain storage of mine waste and its subsequent remobilization.

Active transformation can also result in a channel metamorphosis from meandering to braidedin response to the large sediment input, followed by incision and reversion to a single channel aftermining (and the sediment supply) has ceased. This is well-illustrated by the River Nent, a tributaryof the River Tyne in north-east England, which was affected by large inputs of lead-, zinc- andcadmium-bearing mining waste in the eighteenth and nineteenth centuries ad. Here, the singlechannel and floodplain before 1820 was transformed into a broad, aggrading valley floor due tothe large influx of fine-grained, metal-rich sediment. After mining ceased in the early twentiethcentury, lateral reworking was initiated, and incision of this floodplain began sometime between1948 and 1976, eventually exposing bottom gravels by 1984 (Case Figure 3.2b).

By contrast, in passive dispersal, the channel and floodplain are not disrupted by the influx ofmine waste, because that waste is carried as part of the ‘normal’ sediment load. This is regardedas a system at equilibrium. ‘Natural’ and contaminated sediments are transported together,deposited in, and remobilized from, lateral and vertical (overbank) accretionary deposits, withlittle physical impact on the system.

Even though the terms active transformation and passive dispersal were defined for mining-affected river systems, they can equally be applied to rivers that are affected by other sediment-borne contaminants such as nutrients, organics and radionuclides.

Relevant reading

Gilbert, G.K. (1917) Hydraulic Mining Debris in the Sierra Nevada. Professional Paper 105, US GeologicalSurvey,Washington.

Graves, W. & Eliab, P. (1977) Sediment Study: Alternative Delta Water Facilities – Peripheral Canal Plan.Central District, California Department of Water Resources, Sacramento.

Lewin, J. & Macklin, M.G. (1987) Metal mining and floodplain sedimentation in Britain. In: InternationalGeomorphology 1986, Part 1 (Ed. V. Gardiner), pp. 1009–27. Wiley, Chichester.

Lewin, J., Bradley, S.B. & Macklin, M.G. (1983) Historical valley alluviation in mid-Wales. Geological Journal18, 331–50.

Miller, J.R. (1997) The role of fluvial geomorphic processes in the dispersal of heavy metals from mine sites.Journal of Geochemical Exploration 58, 101–18.

ESC03 9/7/06 3:16 PM Page 100


Tributary streams

Active gravel area

Terrace

1984

1980

1976

1948

1899

1861

1820

1775

0 500metres

N

Yuba River(Narrows near Smartsville)

Yuba River (Marysville)

Sacramento River (Sacramento)

Year

Year

Low

Wat

er E

leva

tion

(Yub

a R

iver

)

Low

Wat

er E

leva

tion

(Sac

ram

ento

Riv

er)

1850 1900 1950

1850 1870 1890 1910

Ele

vatio

nE

leva

tion 60

50

40

10

0

30

20

10

0

60

50

40

30

Yuba River

SacramentoRiver

Ele

vatio

n 260

250

240

(i)

(ii)

(a) (b)

Case Fig. 3.2 (a) Changes in low-flow channel bed elevations along the Yuba and Sacramento Rivers, California: (i) Gilbert’s(1917) original data (modified from Gilbert 1917); (ii) additional work by Graves & Eliab (1977) that uses and extends Gilbert’sdata (modified from Graves & Eliab 1977). (Figures adapted from Miller 1997.) The plots illustrate active transformation of therivers by channel bed aggradation and subsequent degradation. (b) Channel changes (plan view) between 1775 and 1984 on the River Nent, north-east England. Active transformation has resulted in a change from a single thread to a braided channel, anddeposition of fine-grained mining waste. Subsequently the latter has been incised since sometime between 1948 and 1976.(After Lewin and Macklin 1987.)

ESC03 9/7/06 3:16 PM Page 101


The residence times of contaminants in flood-plains can be very long, ranging from 101 to 103 years. This was illustrated by Helgen &Moore (1996), who estimated that it would take several thousand years to completely remobilize metal-bearing tailings deposited onthe floodplain of the Clark Fork River flood-plain, Montana, USA. The length of time thecontaminants remained stored in the floodplainenvironment generally depends on the rates ofpost-depositional physical, biological and chem-ical remobilization (see below), and on the geomorphology of the basin or reach. In stablereaches (e.g. floodplains far from the activechannel, valley reaches with low stream power)or reaches where aggradation rates are high,contaminants can be stored for extremely longperiods (James 1989; Miller 1997), whereas inmore active reaches (e.g. near-channel flood-plains, valley reaches with high stream power),the storage period may be short (James 1989).

Lewin & Macklin (1987) showed that thedeposition and storage patterns of sediment-associated metals can be related directly to flood-plain geomorphological processes and channelsedimentation styles. In laterally mobile braided

and meandering river environments, aggradationof contaminated sediments occurs in floodplainsand bars in reworked valley floors (Wolfenden& Lewin 1977; Lewin et al. 1983; Macklin &Lewin 1989). The patterns of sediment metalconcentrations vary considerably across flood-plains in these environments, according to theages and quantities of contaminated sedimentsdeposited (Lewin et al. 1983; Macklin & Lewin1989).

Floodplain morphology often affects thenature of contaminated sediment deposition. In reaches or rivers that experience little lateralmovement, contaminated sediments are addedto floodplains by overbank deposition as a thin veneer over the whole floodplain surface(Lewin & Macklin 1987; Bradley & Cox 1990;Chesnokov et al. 2000) (see Case Study 3.3).River-bed incision at the end of the eighteenthcentury in the lower South Tyne valley, England,restricted metal-contaminated overbank sedimentdeposition to a relatively narrow trench adjacentto the contemporary river (Macklin 1988). Inthe River Aire, England, the construction of floodembankment structures at Beal confined andaccelerated sedimentation in the near-channel

Case Study 3.3 Styles of deposition of sediment-borne radionuclide contaminants in the RiverTecha, Urals, Russia

The River Techa in the Chelyabinsk region of the Urals in Russia is typical of river systems affectedby releases from nuclear facilities. Liquid radioactive wastes with a total activity of 1017 Bq, ofwhich 12.2% was the radiogenic isotope caesium-137 (137Cs), were released to the River Techafrom the Mayak plutonium facility between 1949 and 1952. These waste releases coincidedwith large floods, resulting in widespread contamination of the Techa floodplain by 137Cs. Waterreservoirs and by-pass channels were constructed between 1951 and 1961 to prevent furtherradionuclide release to the river system, but the contaminated floodplain is currently exploitedfor fishing, bathing, pasture and hay collection by residents of the village of Muslomovo.

Subsequent study showed that much of the liquid waste had precipitated or sorbed onto sediment grains, with the total deposition of 137Cs within a surveyed area of 2.5 km2 estimatedat 6.6 TBq. Two major styles of deposition of 137Cs-contaminated sediment occurred. In manyareas, deposition of highly contaminated sediment was restricted to a narrow zone near theriver (Case Fig. 3.3a). This zone is confined by steep river banks (gradient of 0.0015) and has a relatively narrow floodplain (up to 75 m); maximum 137Cs deposition above 7.5 MBq m−2 islocalized in areas of bank height up to 1 m above normal water level (Case Fig. 3.3b). Further

ESC03 9/7/06 3:16 PM Page 102


away from the river, 137Cs concentrations decline significantly, generally within 15–20 m fromthe river (Case Fig. 3.3c). Even though these depositional sites concentrate highly contaminatedsediment, they only constitute a small percentage of the total 137Cs within the surveyed area andare thus referred to as ‘transit zones’.

A large proportion of 137Cs-contaminated sediment is deposited in ‘sedimentation zones’that comprise a well-drained floodplain with large meanders to the east of the Muslomovo anda marshy floodplain (with a gradient of 0.0003 and width of 700 m) to the west of the village.The latter contains more than 2 TBq of 137Cs. The total deposition of 137Cs-contaminated sediment per unit length of the River Techa corresponds well to the contaminated floodplainwidth distribution along the river (Case Fig. 3.3b), showing that the floodplain is a very effectivecontaminant sink, and that low-lying, wide floodplain areas such as those discussed above storelarge amounts of 137Cs.

Relevant reading

Chesnokov, A.V., Govorun, A.P., Linnik, V.G., et al. (2000) 137Cs contamination of the Techa river flood plainnear the village of Muslumovo. Journal of Environmental Radioactivity 50, 179–91.

Trapeznikov, A.V., Pozolotina, V.N., Chebotina, M.Ta., et al. (1993) Radioactive contamination of the Techariver, the Urals. Health Physics 65, 481–8.

R. Techa

MBq/m2

10521

0.05

G, MBq/m2 H, m

0 5 10 15 20 25

12

10

8

6

4

2

0

2.2

2.0

1.6

1.2

0.8

0.4

0

Y, m

12

100

200

300

400

W, m

Q (x)W (x)

0 4 8 12 16

X, km

0

300

600

900

1200

Q, GBq/km

(a)

(b)

(c)

100 m

Case Fig. 3.3 (a) Caesium-137 distribution on the Techa floodplain. Note the high concentration of 137Cs-contaminated sediment close to the channel. (b) Caesium-137 deposition (G) and bank height (H) abovewater level as a function of the distance from the channel (Y).Deposition in this ‘transit zone’ area is confined to the near-channel area by the steep valley sides. (c) Caesium-137contaminant distribution along the River Techa, where Q isthe total 137Cs deposition per unit river length and W is thewidth of the contaminated floodplain. (Adapted fromChesnokov et al. 2000.)

ESC03 9/7/06 3:16 PM Page 103


zone several metres above the adjacent valley floor(Hudson-Edwards et al. 1999a). Contaminantsassociated with fine-grained sediments also pre-ferentially accumulate in topographically lowareas on floodplains (Wolfenden & Lewin 1977;Bradley & Cox 1990). In floodplains affected by overbank deposition, sediment contaminantconcentrations are either highest immediatelyadjacent to the active channel, dropping sharplywith increasing distance from the river (Macklin1988; Leigh 1997; Chesnokov et al. 2000; seeCase Study 3.3), or are more uniform across thefloodplain (Bradley & Cox 1990). Macklin (1996)suggested that these differences were due to grain-size controls on contaminant concentrations.

3.4.2.4 Uses of overbank sediment profiles

Overbank sediment profiles have many uses in contaminated river systems. Because over-bank sediments record both the natural andanthropogenic geochemical evolution of flood-plains, one of their main uses is as sampling andmapping media to assess metal contaminationcaused by mining and industrial activity (Lewin& Macklin 1987; Macklin et al. 1992, 1994;Miller 1997; Hudson-Edwards et al. 1999a).This is generally carried out by comparing near-surface overbank contaminant concentrationsto those in sediments buried deeper in the sediment profile. Macklin et al. (1994) stressedthat this type of assessment should be carriedout only in conjunction with geomorphologicalmapping and dating of representative floodplainoverbank profiles at a number of reaches, totake into account any lateral variations in metalconcentrations and to cover as wide an age rangeas possible. Metal concentrations and ratios in overbank sediments also have been used asstratigraphical markers for provenancing (e.g.Passmore & Macklin 1994), dating (Davies &Lewin 1974; Lewin et al. 1977; Macklin & Lewin1989; Macklin et al. 1992) and examining thecontaminant sedimentation histories of verticallyaccreted fine-grained overbank deposits (Swennenet al. 1994; Hudson-Edwards et al. 1999a). Thisis possible because metal concentrations andratios vary systematically with respect to the

contaminant inputs from mining or industrialareas (e.g. Knox 1987; Swennen et al. 1994).

3.4.2.5 Contaminant remobilization

Contaminants stored on floodplains may beremobilized through chemical processes (changesin redox and pH; Hudson-Edwards et al. 1998)and through physical erosion, as shown by Brunet& Astin (2000), who reported that elevated dis-charges of inorganic N and P were associatedwith increased autumn rainfall and sediment con-veyance in the River Adour, south-west France.Physical remobilization may take place long afterthe primary contaminating activity (e.g. mining)has ceased (Miller et al. 1998). In the CarsonRiver basin, west-central Nevada, USA, Milleret al. (1998) demonstrated that lateral instabil-ity, coupled with channel-bed incision, resultedin the exposure and erosion of mining-related,Hg-contaminated sediment from bank sediment,and suggested that the valley fill was the prim-ary contemporary source of Hg to the river.Contaminant remobilization is often initiatedby natural (climate) or anthropogenic changes(land use) that cause modifications to sedimentload and delivery, and ultimately erosion anddeposition (Lewin & Macklin 1987; Macklin &Lewin 1989; Macklin 1996; Miller 1997).

3.5 MANAGEMENT AND RESTORATION OF

FLUVIAL SYSTEMS

River systems are continually being brought undermanagement owing to human requirements forwater, agriculture, industry and urbanization.Indeed, there are few rivers in the world todaythat are not regulated or managed to some degree.River management is carried out for a large num-ber of reasons, based on perspectives that riversare either potential hazards or benefits toanthropogenic activity (Table 3.4).

In many cases, river engineering and man-agement has not been carried out with a fullunderstanding of fluvial geomorphology. Gilvear(1999) has identified fundamental areas wherefluvial geomorphologists can contribute to the

ESC03 9/7/06 3:16 PM Page 104


engineering of rivers and floodplains, and sug-gested that, ideally, river management schemesshould incorporate all of these factors from theplanning stage onwards. The first of these is thatlateral, vertical and downstream connectivityand relationships between planform, profile andcross-section should be identified. In other words,schemes should consider the whole river basinrather than only local reaches. This premise isbased on the principles that river channels arethree-dimensional, with longitudinal, transverseand vertical dimensions that are modified inresponse to changes in water and sediment fluxes.The interconnectivity between different parts of river systems means that change in one partwill eventually result in change within a con-tiguous part (Knighton 1998). This is clearlyshown in the cases of dam construction, wheretrapping of the upstream sediment load leads to bed degradation and loss of habitat down-stream (Kondolf 1995; section 3.4.1.3). It isthus important for engineers to view rivers frombasin, rather than reach, perspectives (Knighton1998) because by doing so they will be able tounderstand the underlying causes for potentialproblems, rather than their local (reach-based)manifestations (Sear et al. 1995).

The second of Gilvear’s (1999) recommenda-tions is that the chronology of events, landformsand sedimentary deposits in basins over a rangeof time-scales be determined for river systems.

James (1999) stressed this time aspect of fluvialchange, because this is highly relevant to thelong-term stability of engineered structures andflood-risk assessments. Gilvear (1999) suggestedthat time-scales of less than 1000 years are the most significant to managed rivers. Severalstudies have demonstrated that river channelsare also extremely sensitive to environmentalchange and can adjust very rapidly, from days,seasons to years (Coulthard et al. 2000). Studiesof fluvial landforms as indicators of stability andinstability, and documentation of sedimentaryhistories within river basins, can be carried outby geomorphologists to aid prediction of futurechange (Gilvear 1999).

Recommendation three follows on from two,in that geomorphology should be linked to envir-onmental change. By collecting field geomor-phological data and combining them with otherfield (e.g. age dating, pollen analysis, vegetationmapping) and historical data (e.g. chronicles,weather diaries; Newson 1992), the fluvial geo-morphologist can identify the processes (e.g. climate change) responsible for river landforms,and predict how rivers may respond to theseprocesses in the immediate future. Climate changeand the related frequency of extreme events suchas floods and related large-scale sediment transfercan result in structural changes to rivers anddamage to built structures (such as channels ordams). Baker (1994) has pointed out that such

Table 3.4 Perspectives and objectives of river management. (Based on Knighton 1998, p. 330.)

River as hazard River as resource

Bank protection AestheticsBridge stability AgricultureDeforestation ConservationFlood control – channelization, dams EcologyFloodplain zonation FisheryLand drainage – agricultural drains, road drainage, Heritageurban stormwater systemsContamination of water and sediment NavigationSoil erosion and sediment transport RecreationGully development Rivers as international- to district-level boundaries

UrbanizationSand and gravel extractionWater resource – irrigation, industrial and municipal supplies, power generation

ESC03 9/7/06 3:16 PM Page 105


extreme events may not be predicted by usingstatistical analysis of hydrological data, becausethey are periodical on a scale outside of theserelatively short-term data. Baker et al. (1983)suggested that careful, combined geomorpho-logical and palaeohydrological studies of palaeo-flood hydrology and sedimentary deposits canhelp to elucidate these patterns, to help predictthe real flood risk and potential hazard, and tomake long-term plans for river systems.

Gilvear’s fourth point is that engineers shouldrecognize the links between landforms and pro-cesses in establishing and controlling fluvialecosystems. Pools, riffles, undercut banks andbackwaters are vital habitats for many riverinespecies. In fact, hydrogeomorphological variableshave been shown to be more important thanvegetation in terms of controlling species popu-lations (Bickerton 1995). Vegetation, in turn, iscontrolled by sediment type, organic matter con-tent and stability, as well as hydraulic factors.Efforts at restoring riverine wetland ecology thatfocus on re-establishing pre-existing landformsby removing unnecessary sediment or addingstructures to encourage the formation of habitat-friendly landforms (Brookes 1992) are generallythe most successful.

As sediment transfer is a key process thatlinks engineering and fluvial geomorphology(Sear et al. 1995), an understanding of thesources, distribution and sinks of sediment isfundamental to engineering and managementpractices. Having the tools to provenance sus-pended sediment in river systems is therefore ofvital importance in controlling erosion-sensitiveareas, monitoring and maintaining water andsediment quality, and developing geomorpho-logical models (Walling 1990). Furthermore,knowledge of the influence of hydraulic factorson channel change, sediment transport processes,sediment budgets and deposition and storage ofsediment, on a variety of time-scales, is necessaryto improve river engineering and reduce hazardssuch as floods (Anthony & Julian 1999).

River management thus ideally involves thecollaboration of engineers and fluvial geomor-phologists, as well as end-users and managers of the river’s hazards and resources. Although

many schemes are focused primarily on the pro-tection, enhancement and effective use of watersupplies, they also cover the important issue of sediment management. Factors include theerosion and sedimentation, reducing contam-ination, restoring habitats (see below) and pro-tecting productive floodplain zones. Becausewater and sediment transfer in river basins areintimately connected, many types of manage-ment involve both. Today, for example, climatechange and other factors are forcing govern-ments to adopt ‘soft’ engineering options forrivers, particularly in the developed world, withthe creation of washlands (involving removal oraddition of sediment) as a means of flood con-trol (Newson 1992).

One branch of river management, river re-storation, has seen increasing growth since itsinception in the 1980s (Gore 1985). Restorationinvolves the creation of sustainable geomorpho-logical features that are favourable habitats for riverine biota to recover from damage or tore-establish (Newson 1992). Methods include construction of habitats, management of riparianzones, restoration of hydrological stability andimprovement of water quality (Newson 1992;Gilvear 1999). Although many schemes involvethe creation of stable landforms (pools, stableriffles), increasingly, the importance of ephemeralforms such as bars, splays and scours is beingrecognized (Marston et al. 1995). Other methods,including revegetation, reduction of channelbanks to reduce erosion and restoration ofmeanders, are also important (Newson 1992).

3.6 FUTURE ISSUES

A large number of issues face fluvial environ-ments in terms of sedimentation and sedimentquality. Some of the most important of these areoutlined below.

3.6.1 Impacts of climate change

The predicted changes in temperature, precipita-tion and other global-warming-related effectson sedimentary systems are outlined in Chapter

ESC03 9/7/06 3:16 PM Page 106


1. During global warming the magnitude andfrequency of streamflow can be significantlychanged, resulting in increased frequency ofhigh-flow events (Macklin 1996). This can resultin increased flooding, especially flash flooding,hazards such as debris flows and landslides, andchannel change. Rumsby & Macklin (1994) haveshown that in northern England since ad 1700,channel incision is related to clustering of largefloods, and lateral reworking and sedimenttransfer occurs during episodes of low flood fre-quency. They suggested that increased floodingin the future (associated with global warming)could therefore result in increased channel ero-sion. Longfield & Macklin (1999) attributedhigh-magnitude, high-frequency floods in theYorkshire Ouse Basin, UK, during the 1980s and1990s, to increases in the frequency and vigourof cyclonic atmospheric circulations; these alsohave been related to significant remobilizationof mining-contaminated sediment (Dennis et al.2003). Increases in flooding are also related to ‘priming’ of land surfaces by agriculturalpractices, which leads to increases in runoff and accelerated sediment delivery to floodplain environments (Knox 2001). Global warmingcould also lead to increases in the magnitude,intensity and frequency of forest fires, which mayalso exacerbate bank erosion, sediment trans-port and widespread alluviation.

Cellular computer models are increasinglybeing used to study the development of riverbasins, and the effects of climate, base level andother environmental changes on erosion andsedimentation processes (Veldkamp & VanDijke 1998; Coulthard et al. 2000; Coulthard &Macklin 2001). These models are most effec-tive when combined with field-based studies(e.g. Coulthard & Macklin 2001). Models suchas these will, in the future, prove invaluable inassessing possible future geomorphological andsedimentological changes in river systems, andlinking these to past, analogous changes. Anexample is the FLUVER model, constructed byVeldkamp & Van Dijke (1998) and applied tothe Allier–Loire basin in France. The model takesinto account tectonic uplift, changes in riverlongitudinal profiles and changes in effective

precipitation and sea (base) level. It showed that,although changes in erosion and sedimentationwere related to climatic changes in some instances,the role of tectonic uplift and changing baselevel had previously been underestimated.

3.6.2 Impacts of increased anthropogenicdisturbance

As requirements for natural resources and landincrease, river systems are, and will continue to be, put under pressure. Demand for newhousing on almost all the world’s continentsmeans that floodplains are increasingly targetedas desirable building sites. This has had drasticconsequences in many areas, especially withrecent increases in large-scale floods associatedwith climate change. In the UK in the winter of2000–2001, for example, severe winter stormsresulted in the flooding of 10,000 properties in over 700 locations and £1 billion worth ofdamage. In terms of floodplain sediments, floodssuch as these cause deposition of potentiallycontaminated sediment in living areas, causing a potential risk to human health. The increas-ing urbanization of floodplains has also led to a loss of accommodation space for sedimentdeposition, resulting in silting of channels andreservoirs. Government organizations such asthe UK Environment Agency have recommendedmore stringent restrictions on building on flood-plains (Environment Agency 2004), in additionto more efficient land management.

Increased river regulation to provide powerand surface water storage will continue to havean impact on sediment supply (as well as riverand estuary ecology) both upstream and down-stream of dams and reservoirs. Dam construc-tion increased dramatically from the 1950s tothe 1970s, when two to three new large damswere being commissioned per day (World Com-mission on Dams 2000). Even though dambuilding has declined since the 1970s, majorlarge dam projects are still ongoing, particularlyin China (which accounts for 46% of damsbuilt), the USA and India (World Commissionon Dams 2000). Knighton (1998) has pointedout that the long-term effects of the explosion

ESC03 9/7/06 3:16 PM Page 107


in twentieth century river regulation projectsmay not be felt for at least another 100 yearsbecause river adjustment is a long-term processes(cf. Petts 1984).

Although mining in many areas of the worldis now regulated and under strict environmentallegislation, there are still areas where miningcontinues to have an impact on sediment andsediment-borne contaminant supplies to rivers.An example of this is the Ok Tedi/Fly River inPapua New Guinea, which, according to 1997figures, received approximately 66 Mt of miningwaste per year from the Ok Tedi Cu–Au mine,causing increases in the Cu- and other metal-bearing suspended sediment load by up to five to ten times natural background (Hettler et al.1997). This is compounded by the steep relief inthe area, which is linked to avalanches and land-slides in the mining area (Hearn 1995). In otherareas, it is not the mining itself, but relatedactivities that threaten river systems. The 1998Aznalcóllar tailings dam spill in southwestSpain (Grimalt et al. 1999) and the 2000 BaieMare tailings dam spills in Romania (Macklin et al. 2003), which released 2 × 106 m3 and40,000 t, respectively, of metal-, arsenic- andcyanide-rich tailings downstream, are notableexamples of this. Events such as these overwhelmthe natural sediment load of river systems, andcause immediate and long-term threats to thehealth of ecosystems and humans. Until saferalternatives are found for the storage of minewaste, events such as these will probably con-tinue to occur regularly (Macklin et al. 2003).

3.6.3 Other impacts

The storage of contaminated sediments in flood-plains has been flagged as a potential serious long-term problem and, indeed, has been described asa ‘chemical time bomb’ (a concept defined byStigliani 1991) by many workers in river systems(Stigliani 1991; Lacerda & Salamons 1992).Although considerable effort is being put intounderstanding the fate of sediment-bound con-taminants (Hudson-Edwards et al. 1998, 2003),the interplay between their physical (river bankand bed erosion, land drainage and development),geochemical (changes in nutrient loading, redoxand pH) and biological remobilization (plantuptake and microbial degradation) is poorlyunderstood (Macklin 1996). Moreover, Macklin(1996) warned that physical remobilization ofthese contaminants is currently increasing owingto global warming. Sediment alone has also beenflagged by many government environmentalbodies as a significant threat to river systems interms of its weight and volume. This has beenshown in this chapter, where processes such as increased alluviation in regulated rivers canhave large secondary effects.

Finally, the present rapid rate of urbanization,particularly in the tropics (Gupta & Ahmad1999), means that rivers are, and will continueto be, put under pressure for their resources and hazard management (cf. Knighton 1998).This requires more effective management thatintegrates good fluvial geomorphological andengineering practice.

ESC03 9/7/06 3:16 PM Page 108

4.1 INTRODUCTION AND BACKGROUND

Lakes are important sedimentological environ-ments. When a river enters into a lake, the sectionarea becomes wider, hence the flow velocity ofthe running water is reduced and the particlessuspended in the water end up in a much calmerenvironment where they can settle out. Thecoarse and heavy particles carried by the river willsettle more rapidly than the smaller, less denseparticles. As a result, lakes function as ‘sedimenttraps’ for materials and pollutants transportedin rivers. The consequences of this are that lakesaccumulate and retain pollutants, that lake sedi-ments may be heavily contaminated and thatlake sediment cores are excellent for studies ofthe historical development of contamination. In addition, lake sediments are good records of temporal changes in the landscape and areimportant systems for studying the ecosystemeffects of pollutants. Lakes are also extremelyimportant freshwater resources, important forrecreation and totally dominate the landscape of certain regions. As a result, there are manyreasons why lakes are target systems for researchand management.

4Lake environments

Lars Håkanson

There is no generally accepted terminology asto what is a lake, a pond or an inland sea, except,of course, that ponds are small and inland seaslarge. From Table 4.1, however, one can note thatthere are 227,000 ponds, small lakes, lakes andlarge lakes in Sweden alone and that ponds aresmaller than 1 ha and objects larger than 100 km2

may be called large lakes. The primary focus ofthis chapter is to discuss the role of sedimentsand suspended particulate matter (SPM) in lakes.How do sediments and SPM influence the sedi-mentological and ecological function of lakesystems? What are the sedimentological controlson the distribution and effects of environmentalpollutants? These could be, and have been, dis-cussed from many perspectives and scales. Thischapter focuses on general principles and pro-cesses, many of which are analogous to thosedescribed in the chapters on riverine, coastal andmarine environments. The format of the chapterwill be to first discuss schemes for lake classifica-tion and controls on lake forms (section 4.1).Section 4.2 discusses sediment sources and theresulting products. Section 4.3 focuses on naturaland anthropogenic disturbances on lake sediment-ary systems; section 4.4 focuses on sediment

Area (km2) Category Number Cumlative number Per cent

> 1000 Large lakes 3 0.0013100–1000 Large lakes 19 19 0.008410–100 Lakes 362 381 0.161–10 Lakes 3987 4368 1.90.1–1 Small lakes 19,374 23,742 23.270.01–0.1 Small lakes 59,500 83,242 36.60.001–0.01 Ponds 144,000 227,000 100

Table 4.1 The number of lakes in Sweden arranged by size (1 km2 = 100 ha). (Data from Monitor1986; Håkanson & Peters 1995.)

ESC04 9/5/06 3:02 PM Page 109

110 LARS HÅKANSON

dating and sediment records; section 4.5 discussesbriefly the role of future climate change from asedimentological perspective; and section 4.6concerns lake management issues.

4.1.1 Lake types and classification

A fundamental question in lake management is:‘if there is a change in an important lake vari-able, such as the phosphorus concentration, theconcentration of suspended particulate matter(SPM) and/or lake pH (which can be related toacid rain), will there also be changes in key func-tional groups of organisms and changes in eco-system function and structure?’ Is it possible toquantify and predict such changes? Lake classi-fication systems may not in themselves provideanswers to such questions, but they can providea scientific framework for such analyses. Differenttypes of lakes have different types of sediments.Such relationships will be discussed in this chap-ter. To start that discussion, this section presentsdifferent approaches to classify lakes.

Table 4.2 gives a compilation (from Hutchinson1957) of all existing lake types on Earth, asclassified according to form-creating processes.Most lakes are of glacial origin, i.e. they wereformed by erosional or depositional glacial activ-ities. In many parts of the world, for example

northern Africa and central Australia, lakes are rare, whereas in other areas they dominate the landscape, for example Sweden (Table 4.1),other parts of Scandinavia and the northernparts of North America and Russia. It also canbe seen from Table 4.1 that there exists a clearrelationship between number and size; there aremany more small lakes than large lakes in theglacial landscape of Sweden.

Many schemes for classifying lakes have beenput forward over the past 50 years (e.g. OECD1982; Nürnberg 1996; Nürnberg & Shaw 1998).In addition to the criteria discussed below, othercriteria to classify lakes have also been used,including the oxygen content in the water andthe species composition of the flora and fauna(Håkanson & Boulion 2002). The aim here is toextract some key aspects of lake classificationschemes in order to provide a framework for thefollowing sections of this chapter.

A trophic level classification system of lakes is given in Table 4.3. This scheme recognizesfour trophic levels: oligotrophic, mesotrophic,eutrophic and hypertrophic lakes. The trophicstatus of a lake is usually estimated by mean values of primary production measured for thegrowing season. Note that there is a significantoverlap between the different trophic categoriesin Table 4.3. For example, in low-productive

Table 4.2 Major lake types on earth according to Hutchinson (1957).

Type

Tectonic lakesVolcanic lakes

Landslide lakesGlacial lakes

Solution lakesFluvial lakes

Aeolian lakesShoreline lakesOrganic lakesAnthropogenic lakesMeteorite lakes

Subtype

Lakes in direct contact with iceGlacial rock basinsMorainic and outwash lakesDrift basins

Plunge-pool lakesFluviatile damsMeander lakes

Example

Basins in faults (like Lake Bajkal and Lake Tanganyika)Maars, caldera lakes and lakes formed by damming oflava flowsLakes held by rockslides, mudflows and screesLakes on or in ice and lakes dammed by iceCirque lakes and fjord lakesLakes created by terminal, recessional or lateral morainesKettle lakes and thermokarst lakesLakes formed in caves by solution

Strath lakes, lateral lakes, delta lakes and meresOxbow lakes and cresentic levee lakesBasins dammed by wind-blown sand and deflation basinsTombolo lakes and spit lakesPhytogenic dams and coral lakesDams and excavations made by manMeteorite craters

ESC04 9/5/06 3:02 PM Page 110

LAKE ENVIRONMENTS 111

(oligotrophic) lakes, the concentrations of totalphosphorus (TP; generally the key nutrient regu-lating primary production in lakes) may varywithin a year from very low to high values.

A similar classification system is shown in Fig. 4.1. It uses values of chlorophyll a (μg L−1)as the basic criteria to classify lake trophic levelbecause chlorophyll (Chl) is relatively easy andinexpensive to measure and it is a generally usedbiological measure of phytoplankton biomass.In oligotrophic systems the mean summer Chl

values generally vary from 0.1 to 1 μg L−1, inmesotrophic systems from 1 up to 10, in eutro-phic systems from 10 to 100, and in hypertrophicsystems the Chl values are generally higher than100 μg L−1. Thus, the boundary Chl values forevery trophic class can be set equal to a powerwith the base 10.

A Trophic State Index (TSI) is defined accordingly:

TSI = 25 · (log(Chl) + 1)

Table 4.3 Characteristic features in lakes of different trophic categories (Modified from OECD 1982; Håkanson & Jansson 1983.)

Trophic level Primary production Secchi Chl-a Algal volume* Total P† Total N† Dominant fish(g C m−2 yr−1) (m) (mg m−3) (g m−3) (mg m−3) (mg m−3)

Oligotrophic < 30 > 5 < 2.5 < 0.8 < 10 < 350 Trout, whitefishMesotrophic 25–60 3–6 2–8 0.5–1.9 8–25 300–500 Whitefish, perchEutrophic 40–200 1–4 6–35 1.2–2.5 20–100 350–600 Perch, roachHypertrophic 130–600 0–2 30–400 2.1–20 > 80 > 600 Roach, bream

*Mean value for the growing period (May–October)†Mean value for the spring circulation.

HSI Colour (mg Pt L−1)

1000100

82

66

48

28

0

300

100

30

10

3

1.1

0

1

25 50 75

Hypertrophic

Ultraoligohumic

Oligo-humic

Meso-humic

Polyhumic

Hyperhumic

EutrophicMesotrophicOligotrophic

100 TSI

10 100 1000 Chlorophyll (µg L−1)

Fig. 4.1 The relationship between lake colour and humic state on the y axis and lake chlorophyll and trophic state on the x axis anddata from 936 lakes: HSI, Humic State Index; TSI, Trophic State Index. (Modified from Håkanson & Boulion 2002.)

ESC04 9/5/06 3:02 PM Page 111

112 LARS HÅKANSON

The TSI varies from 0 to 100. A TSI value of 100relates to a Chl value of 1000 μg L−1 (the limit forhypertrophic lakes). A TSI value of 75 relates to a Chl value of 100 μg L−1, the limit betweeneutrophic and hypertrophic lakes, etc. At Chlvalues < 0.1, then the TSI should be set to 0.

In a similar manner to trophic status, lake humicstatus can be determined using lake colour valuesas a criterion for humic state. There are severalreasons to base the humic state index on colour.Coloured substances such as humic and fulvicacids reduce light transmission through water,affect bioavailability and ecological effects ofnutrients and contaminants, affect plankton meta-bolism and may inhibit secondary production(Wetzel 2001). The colour value is generally deter-mined by the comparative method using coloureddiscs and expressed in mg Pt L−1 (Håkanson et al.1990). In this way, a Humic State Index (HSI)may be defined on a scale from HSI = 100, hyper-humic, to HSI = 0, ultraoligohumic, where: HSI =(100/3)log(colour − 3). For example, for HSI =100 and colour = 300 (mg Pt L−1) gives HSI = 82,etc. (Note that if colour < 3 (mg Pt L−1), thenHSI should be set to 0; but colour values < 3 and> 1000 mg Pt L−1 are very rare.)

The lake classification scheme in Fig. 4.1 usescategories of primary production (autotrophy)on the x axis and categories of catchment influ-ence from soil, bedrock, vegetation and land-useactivities (allotrophy) on the y axis. This systemto classify lakes has much in common with the

classification scheme for lake sediments (describedlater in Fig. 4.3). Lakes also can be classifiedaccording to their thermal properties, where thelake water is separated into surface water (epi-limnetic water) and deep water (hypolimneticwater) by a thermocline (temperature stratifica-tion), i.e. warmer, lighter waters are found on topof colder, denser waters. Temperature stratifica-tions occur in dimictic lakes (i.e. lakes that mixtwice a year, generally during spring and autumn).Lakes also can be monomictic (such lakes circu-late once a year) or polymictic (which mix manytimes per year; see Wetzel 2001).

4.1.2 Controls on lake form

The size and form of lakes regulate many gen-eral transport processes, such as sedimentation,resuspension, diffusion, mixing, burial andoutflow (Fig. 4.2), which in turn regulate manyabiotic state variables, such as concentrations ofphosphorus, suspended particulate matter, pHand many water chemical variables, colour andwater clarity. These in turn regulate primaryproduction, which in turn regulates secondaryproduction, for example of zooplankton andfish (see Håkanson 2004). The morphometry ofa lake depends on the nature of the surroundingland area, drainage basin characteristics and theorigin of the lake (see Table 4.2).

From the bathymetric map of a lake, one can define a set of morphometric parameters

Wave base

InflowOutflow

Point source emissions Precipitation

Primary production Mixing ET-sedimentsResuspensionSurface water

= epilimnion

Deep water= hypolimnion

DiffusionSedimentation

Compaction

Burial

Active A-sediments

Biopassive A-sediments (geosphere)

Bioturbation

Fig. 4.2 Illustration of general and fundamental transport processes to, within and from lakes. The major sources of suspendedparticulate matter are the inflow from tributaries, primary production in lakes, point-source emissions and direct deposition of particleson the lake surface. The internal processes (sedimentation, resuspension, mixing, mineralization, biouptake and burial) redistribute thesuspended particles and regulate the retention and outflow of particles from the lake. See section 4.2.2.2 for an explanation of A- and ET-sediments.

ESC04 9/5/06 3:02 PM Page 112


describing lake size, form and specific features.Among these parameters, three main groups canbe identified.1 Size parameters: (i) different parameters inlength units, such as maximum length, waterdepth, shoreline length, maximum depth andmaximum breadth, (ii) parameters expressed inarea units, e.g. two-dimensional water surfacearea and three-dimensional bottom area, and(iii) parameters expressed in volume units, suchas water volume and surface-water volume(epilimnetic volume).2 Form parameters: based on size parameters,for example mean depth and shore development(= shoreline irregularity).3 Special parameters: such as the dynamic ratio(DR = √A/Dm, where A is lake area in squarekilometres and Dm is the mean depth in metres)and effective fetch (a measure in kilometres ofthe free water surface over which the winds caninfluence the waves and hence also the sediments).

4.1.3 Lakes as sedimentary environments

The main sedimentary environments that occurwithin lakes are:1 Beach and nearshore areas dominated by waveprocesses, on- and offshore movement of sedi-ments and nearshore currents (see Teller 2001).The sediments in these areas are generally coarse-grained (sand, gravel, etc.).2 Deltaic areas in lakes with relatively largetributaries transporting coarse silt and sand.3 Shallow areas above the wave base, wherewinds and waves may exert an influence on sedi-ments (the surface-water area approximates to the area above the mean thermocline). Suchsediments are generally well oxygenated andmay be resuspended by storms.4 Deep-water areas occur beneath the wave base(below the thermocline). These areas provide arelatively calm, less well oxygenated sedimento-logical environment for fine materials to settle.5 On slopes inclined at more than 4 m per 100 m length, slope processes dominate the sedi-mentological conditions on lake bottoms. Insuch areas turbidites may form and they caninfluence the sediment record over large areas.

This is an important process with relevance to studies of sediments as historical archives,because turbidity currents can cause older, pre-viously deposited material to settle on top ofmore recently deposited sediment.

4.2 SEDIMENT SOURCES AND SEDIMENT ACCUMULATION

PROCESSES

4.2.1 Sources and characteristics of lake sediments

Five major sources of matter forming lake sedi-ments can be recognized:1 Allochthonous materials (sometimes calledlithogenous materials) – particles and aggregatestransported from land to lakes by rivers (e.g.sand, silt and clay).2 Autochthonous materials (or biogenousmaterials) – sediments produced in the lake byorganisms living in the lake (e.g. silica frustulesor dead phytoplankton).3 Hydrogenous sediments – sediments that arealso produced in the lake but which emanatefrom materials precipitated out of solution, forexample, carbonates (e.g. Pedley 1990; Pedley et al. 1996) and evaporites (e.g. Trichet et al.2001).4 Wet and dry deposition of matter on the lake surface – this is generally a relatively smallcontributor to the total amount of suspendedmatter found in lakes.5 Direct point source emissions of matter, forexample from urban areas and industries.The relative contributions of each source can beassessed for a lake by means of mass-balancecalculations (Fig. 4.2), where the flux (which isequivalent to the flow or transport) from eachsource is calculated. This can be calculated on a monthly basis (g month−1) to obtain seasonalvariations, or annually for overall budgets inorder to rank the relative importance of each flux.The relative contribution of different sourcesdiffers between lakes as a result of variations incharacteristics, lake morphometry and climato-logical regime.

The sediment types within lakes can be clas-sified in a number of ways. One system, widely

ESC04 9/5/06 3:02 PM Page 113

114 LARS HÅKANSON

used, is based on the relationship between theorganic content (estimated by determining loss onignition) and the ratio between the carbon con-tent and the nitrogen content of lake sediments(Fig. 4.3). This sedimentological classificationsystem has much in common with the informa-tion given in Fig. 4.1 for lake classifications.Three main types of matter in lake sediments canbe recognized and each has different chemicalcharacteristics (see Hansen 1961; Håkanson &Jansson 1983).1 The average composition of planktonic mater-ials (autochthonous organic matter) is C106N16P,which gives a C/N ratio of about 5.6, and the losson ignition (LOI) is usually less than 20% dryweight (dw). This group also includes hydro-genous materials precipitated out of solution andbiogenous materials (from organisms).2 Allochthonous organic matter (e.g. humus)generally (see Gjessing 1976; Thurman 1985) hasa C/N ratio of 10–20 and a LOI of > 20% dw.3 Minerogenic (inorganic) matter (such as sandand silt) generally has a C/N ratio of 15–25 anda LOI of < 20% dw.

Further classification schemes related to lakesediments include those based on (i) grain size(gravel, sand, silt or clay; see e.g. Friedman &Sanders 1978), (ii) form-creating processes (such

as shore sediments, glacial deposits, turbiditesfrom turbidity currents, delta sediments, mixedbioturbated sediments or annually layered, laminated sediments; see e.g. Sly 1978), (iii) sys-tems based on sediment colour and/or specificcharacteristics (such as black sediments as areflection of anoxic conditions, brown sedimentsreflecting oxic conditions, light grey calcareousdeposits, greenish grey siliceous deposits, greysediments rich in fibre, very soft brownish sedi-ments, such as gyttja (sedimentary peat consistingmainly of plant and animal residues precipitatedfrom standing water) and dy (finely divided,partly decomposed organic material), and blackmanganese nodules; see e.g. Håkanson & Jansson1983), (iv) systems based on chemical properties(see Table 4.4), or (v) systems that focus on thegeological rather than the geographical origin of the sediments and distinguish between, forexample, glacial, fluvial, post-glacial or aeoliandeposits.

4.2.1.1 Suspended particulate matter (SPM) in lakes

The total amount of any substance X in the wateris often separated into a particulate phase, theonly phase subject to gravitational sedimenta-tion, and a dissolved phase, generally the most

Organic content of lake sediments (LOI, % dw; LOI − 2 . C)

50

40

30

20

10

0

0 5 10 15 20 25 C/N ratio oflake sedimentsOligotrophic

lakesEutrophic

lakes

Autotrophy

Polyhumicdystrophiclakes

Oligohumiclakes

Allo

tro

ph

y

Humic materials:C/N: 10–20

Minerogenicmatter(sand–silt):C/N: 15–25

Plankton:C/N: 5.6

Fig. 4.3 Lake classification from therelationship between the C/N ratio and the loss on ignition (LOI) of surficialsediments. (Modified from Håkanson1995.)

ESC04 9/5/06 3:02 PM Page 114


important phase for direct biological uptake.Operationally, the limit between the particulatephase and the dissolved phase is generally deter-mined by means of filtration, using a pore size of 0.45 μm. Evidently, this is an operational

approach, and many colloidal particles will passthrough such filters.

A general outline of particles in lake water,their origin, standard abbreviations and a classi-fication scheme are given in Fig. 4.4. As stressed,

Table 4.4 A geochemical classification (Berner 1981) of sedimentary environments. C is the concentration (moles L−1). H2S is total sulphide.

Environment

I. Oxic (CO2 ≥ 10−6)II. Anoxic (CO2 < 10−6)

A. Sulphidic (CH2S ≥ 10−6)B. Non-sulphidic (CH2S < 10−6)

1. Post-oxic2. Methanic

Characteristic phases

Haematite, goethite, MnO2-type minerals; no organic matter

Pyrite, marcasite, rhodochrosite, alabandite; organic matterGlauconite and other Fe2+–Fe3+ silicates (also siderite, vivianite, rhodochrosite); no sulphide minerals; minor organic matterSiderite, vivianite, rhodochrosite; earlier formed sulphide minerals; organic matter

Allochthonous material

- Minerogenic particles (clays,silt, sand, Fe–Mn oxides andhydroxides, etc.)- Organic particles (humic matter, living and dead plankton, etc.)

Autochthonous material

- Organic particles (living anddead plankton, etc.)- Inorganic matter (e.g. ashes,shells)

Origin,primary

Origin,secondary

Resuspended particles

Suspended particulate matter(SPM = seston)- Dead versus living- Inorganic versus organic

Total organicmatter (TOM)100%

Particulateinorganicmatter (PIM)

Dissolvedorganic matter(DOM)80%

Particulateorganic matter(POM)20%

Plankton4%

Detritus16%

Conservativeorganic matter68%

Reactiveorganicmatter12%

Particles in lake water

Fig. 4.4 Classificationscheme andnomenclature forparticles in lake water(see also Dubko 1985;Ostapenia 1985).

ESC04 9/5/06 3:02 PM Page 115

116 LARS HÅKANSON

the first differentiation concerns the origin of theparticles: allochthonous particles are transportedto the lake from outside the lake, generally fromtributaries or precipitation onto the lake surface– such particles may be categorized into mineral(or minerogenic) and organic (or organogenic) – and autochthonous particles, which are pro-duced in the lake. These particles can settle on tothe lake bed and some of those particles may beresuspended. This means that the total amountof suspended particulate matter (SPM) found inlakes is generally a complex mix of substancesof different origins with different properties (size,form, density, specific surface, capacity to bindpollutants, etc.).

The SPM may be divided into an organic fraction (POM) and an inorganic one (PIM, par-ticulate inorganic materials; see Fig. 4.4). Totalorganic matter (TOM) is generally divided intoparticulate (POM) and dissolved (DOM) frac-tions. Normally, POM is about 20% of TOM,but this certainly varies between lakes and withinlakes during the year. Normally, about 4% ofPOM is living matter and the rest is dead organicmatter (detritus). About 80% of TOM is gener-ally in a dissolved phase, and of this about 70%is conservative in the sense that it does not changein response to chemical and biological reactionsin the water mass.

4.2.2 Controls on lake sediment transport andaccumulation

The sedimentological conditions in a lake willinfluence almost all processes in the aquatic eco-system. For example, resuspension, especially inlarge and shallow lakes, controls the concentrationof suspended particles in the water, influencingwater clarity and hence the depth of the photiczone, often operationally defined as the Secchidepth, i.e. the depth where a black and white discis lost from eye sight on lowering through thewater (Håkanson & Peters 1995). Resuspension,therefore, has an impact on primary and second-ary production. The production of zoobenthosis also controlled by sedimentological conditions,with high production in oxic sediments and lowproduction in anoxic sediments. The macrophytes

are rooted in the sediments (Fig. 4.5) and, there-fore, sediment conditions control their produc-tion. If, finally, the sediments are contaminated,this could increase the concentrations of harm-ful substances in zoobenthos, and, hence, also infish that eat zoobenthos (benthivores).

Within a lake several processes regulate andcontrol the pathways of both sediments andcontaminants (see Fig. 4.2):1 sedimentation – the transport of matter fromwater to sediments;2 resuspension – the transport of matter fromsediments back to water;3 diffusion – the transport of dissolved sub-stances from sediments back to water;4 mineralization – the bacterial decompositionof organic matter;5 mixing – the upward and downward trans-port of matter;6 bioturbation – the mixing of the depositedmaterials from the movement of the bottomfauna (= zoobenthos – from their eating, diggingand foraging activities);7 compaction – the vertical change in sedimentwater content and sediment density due to theweight of overlying sediments;8 burial, i.e. the transport from biologically activesediments to biopassive (geological) sediments.This latter is transport from the biosphere to thegeosphere of substances that may emanate fromthe technosphere. These processes are discussedin more detail below.

4.2.2.1 Transport, sedimentation and resuspension

The processes of sedimentation, burial and resus-pension are interlinked and to understand themrequires an understanding of bottom dynamicconditions within a lake. In defining the bottomdynamic conditions (erosion, transportation andaccumulation), the following definitions areoften used (from Håkanson 1977).1 Areas of erosion (E) prevail where there is noapparent deposition of fine materials but rathera removal of such materials, for example in shallow areas or on slopes; E areas are generallyhard and consist of sand, gravel, consolidatedclays and/or rocks.

ESC04 9/5/06 3:02 PM Page 116


2 Areas of transportation (T) prevail where fine materials are deposited periodically (areasof mixed sediments). This bottom type generallydominates where wind/wave action regulatesthe bottom dynamic conditions (see Fig. 4.6). It is sometimes difficult in practice to separateareas of erosion from areas of transportation.3 Areas of accumulation (A) prevail where the fine materials are deposited continuously(soft bottom areas). Owing to their fine-grainednature these are the areas (the ‘end stations’)where high concentrations of pollutants mayappear (see Table 4.5).

The water content, grain size and/or the com-position of the material are often used as cri-teria to distinguish different sediment types (seeTable 4.5; or Sly 1978). From the basic Stokes’equation for settling particles (see Chapter 1), as well as for convenience, the limit between

coarse and fine materials can be set at a particlesize of medium silt (0.06 mm). The generallysandy sediments within the areas of erosion andtransport (ET) often have a low water content,low organic content and low concentrations of nutrients, low benthic biomass and few contaminants (see Table 4.5 and Fig. 4.7). Theconditions within the T areas are, for naturalreasons, variable, especially for the most mobilesubstances, such as phosphorus, manganese and iron, which may react rapidly to alterationsin the sediment chemical ‘climate’ (as given e.g. by the redox potential). Fine materials maybe deposited for long periods during stagnantweather conditions. In connection with a stormor a mass movement on a slope, this materialmay be resuspended and transported, generallyin the direction towards the A areas in the deeperparts, where continuous deposition occurs. Thus,

Water surface

Thermocline

Pelagic zone;profundal sediments

Macrophytes- Free-floating- Permanent stands, e.g.(1) emergent with greenparts above the watersurface,(2) floating-leaved, or(3) submerged, with all partsbelow the water surface

Littoral zone;littoral sediments

Zoobenthos, benthic algae and macrophytes

- Size criteria: macro-, meso- and meio-benthos- Feeding criteria:ShreddersCollectorsScrapersPredatorsFilter Feeders

Benthic algae

Epipelic(grows on/insediments)

Epilithic(grows on rocks/stones)

Epiphytic(grows on plantsor animals)

Periphyton(grows on artificialsubstrates, e.g. boats)

Substrate

Fig. 4.5 Compilation of conceptsrelated to sediment-living organisms(zoobenthos, benthic algae andmacrophytes) (see also Vollenweider1968, 1976; Cummings 1973;Brinkhurst 1974; Wetzel 2001).

ESC04 9/5/06 3:02 PM Page 117

118 LARS HÅKANSON

resuspension is a natural phenomenon on Tareas. It should be stressed that fine materialsare rarely deposited as a result of simple verticalsettling in natural aquatic environments. Thehorizontal velocity component in lake water isgenerally at least ten times larger, sometimes upto 10,000 times larger, than the vertical com-ponent for fine materials or flocs which settle

according to Stokes’ law (see Bloesch & Burns1980; Bloesch & Uehlinger 1986).

Resuspension is the physical (advective) trans-port of matter from sediments back to water andmixing is the upward and downward transportof dissolved and suspended particulate matteracross the thermocline (the thermocline is the zonein the water that separates the warmer, lighter

Effective fetch (km)

1 2 4 8 16 32 64 128 256

0

5

10

15

20

25

30

35

40

45

50

Wat

er d

epth

(m

)

Erosion (E)

DET=(30.4.EF)/(EF+34.2)

Transportation (T)

DTA=(45.7.EF)/(EF+21.4)Accumulation (A)

Fig. 4.6 The erosion, transport andaccumulation (ETA) diagram showing therelationship between the effective fetch (EF, i.e. the free water surface over whichwinds influence waves), the water depthand the potential bottom dynamicconditions for individual sites in lakes. DTA is the wave base (WB), i.e. the water depth separating T and A areas. DTA can be predicted from the givenequation. The mean EF value for an entirelake can be approximated by √Area.(Modified from Håkanson 1999.)

Table 4.5 The relationship between bottom dynamic conditions (erosion, transportation and accumulation) and the physical,chemical and biological character of the surficial sediments of Lake Lilla Ullevi Bay (in Lake Mälaren), Sweden. Mean values andcoefficients of variation (CV) in parentheses: n = number of analyses. (Raw data from Ryding & Borg 1973.)

Category

Physical parameters

Nutrients (mg g−1 dw)

Benthic biomass (mg ww m−2)Chemically mobile elements (see also P) (mg g−1 dw)Metals (μg g−1 dw)

Characteristic

Water depth (m)Water content (% ww)Bulk density (g cm−3)Organic content (loss on ignition, % dw)NitrogenPhosphorusCarbon

IronManganeseZincCopperNickel

Erosion (n = 15)

13.0 (0.41)32.6 (0.28)

1.71 (0.087)4.6 (0.48)

0.6 (0.67)0.8 (0.50)0.5 (1.0)

1000–200024.6 (0.42)

0.8 (1.0)41 (0.46)18 (0.50)23 (0.35)

Transportation (n = 10)

17.5 (0.31)67.4 (0.14)

1.26 (0.079)10.7 (0.43)

3.4 (0.35)2.8 (0.75)

22.7 (0.74)3000–4000

53.5 (0.27)3.5 (0.74)

111 (0.24)31 (0.42)40 (0.20)

Accumulation (n = 14)

31.6 (0.25)94.1 (0.024)

1.03 (0.019)24.3 (0.10)

10.7 (0.14)1.6 (0.31)

10.4 (0.16)6000–7000

41.3 (0.077)2.5 (0.60)

189 (0.090)59 (0.10)57 (0.18)

ESC04 9/5/06 3:03 PM Page 118


surface water from the colder, heavier deepwater). Some lakes are constantly mixed and donot develop a thermocline, but most lakes, andcertainly those in northern and boreal landscapes,develop a thermocline in the summer. The sur-face water (the epilimnion) is fundamental for theprimary production of matter. Many importantsedimentological processes take place in thedeep-water zone (= hypolimnion).

There are some basic rules regulating sedi-mentation in lakes (e.g. Thomas et al. 1976;Golterman et al. 1983; Håkanson & Jansson1983; Colman et al. 2000; and Fig. 4.8).1 River action dominates the sedimentologicalproperties in river-mouth areas, where deltasmay be formed if the amount of sandy materialscarried by the tributaries is large enough. Withinthese areas, sedimentation rates generally decreasewith distance from the mouth, and so does thegrain size of the settling particles.2 In open-water areas, dominated by wind/wave

action, sedimentation rates generally increasefrom the wave base to the deepest parts of thelakes. The coarsest materials (sand, gravel) areoften found in shallow waters.3 Current action can dominate in certain areas,such as in narrow straits and along the shoreline.Then the ‘Hjulström-curve’ (see Fig. 1.5) givesthe relationship between critical erosion andcritical deposition of materials.4 Slope-induced (gravity) turbidity currentsappear on bottoms inclined more than about4–5% (Håkanson 1977), and bioturbation gen-erally prevails in oxic sediments (see Table 4.4),where the macro- and meiofauna cause a mixingof the sediments.

4.2.2.2 Determination of bottom dynamic conditions in lakes

The following processes influence internal load-ing and bottom dynamic (E, T and A) conditions

SchiersteinKostheimKolnDusseldorf

100

90

80

70

60

50

40

30

20

10

0

Cad

miu

m (

ppm

)

>2.

0 m

m

1.0

–2.0

0.5

–1.0

0.25

–0.

5

0.12

5–

0.25

0.12

5 m

m–

63 µ

m

20–

63

6.3–

20

2–

6.3

0.63

–2

0.2–

0.63

0.06

3–0.

2

<0.

63 µ

m

Fig. 4.7 Grain-size dependencies of cadmium concentrations in sediment samples from German rivers. (Modified from Förstner &Salomons 1981.)

ESC04 9/5/06 3:03 PM Page 119

120 LARS HÅKANSON

in lakes (Håkanson & Jansson 1983): (i) anenergy factor related to the effective fetch andthe wave base (see the ETA diagram in Fig. 4.6);(ii) a form factor related to the percentage of the lake bed above the wave base (see Fig. 4.9);and (iii) a lake slope factor related to the fact that slope-induced transportation (turbiditycurrents) may appear on bottoms inclining morethan 4–5%.

One approach to calculate the areas whereresuspension occurs (the ET areas) is based onthe wave base and the form of the lake (the formfactor Vd = 3 · Dm/Dmax; Dmax = the maximumdepth, as illustrated in Fig. 4.9). The wave base (WB), which is set equal to the depth, DTA,separating T areas and A areas is given by:

DTA = (45.7 · √Area)/(21.4 + √Area)

An evident boundary condition for this approachis that if DTA > Dmax then DTA = Dmax.

The area above the wave base (AWB = Area −AreaA) may be calculated from the hypsographic

form of the lake, which in turn is calculatedfrom the form factor (= volume development,Vd). An equation expressing AWB, as a functionof Vd, the area of the lake (Area in m2) and themaximum depth of the lake (Dmax in m), is alsogiven in Fig. 4.9. The ET areas are generallylarger than 15% of the lake area because there is always a shore zone dominated by wind/waveactivities, at least in all lakes larger than approx-imately 1 ha. One can generally also in shallowlakes find sheltered areas and deep holes withmore or less continuous sedimentation, i.e.areas that actually function as A areas, so theupper boundary limit for ET is often set at 99%of the lake area.

4.2.2.3 Post-depositional processes

4.2.2.3.1 Bioturbation Bioturbation is the mixingof the deposited materials from the movement(eating, digging and foraging activities) of thebottom fauna (zoobenthos). The sediment clas-sification scheme in Table 4.4 focuses on the

River

River actionWind/wave actionand morphology

Bioproduction

‘pelagic’ sedimentation

Bottom dynamics

wave base Delta sedimentation

River plumesedimentation

123

surface currents

under currents

Turbiditycurrents

Turbiditycurrents

Slope > 5%

Wind (speed, duration,direction and fetch)

Resuspension0123

ErosionTransportation

Accumulation0 1 23

Transportation

Accumulation

1. Erosion (winnowing) – coarse deposits2. Transportation – coarse and fine deposits3. Accumulation – fine deposits (= "pelagic" sedimentation)4. Turbidific sedimentation – fine deposits

1. Delta sedimentation – coarse deposits2. River plume sedimentation – coarse and fine deposits3. Turbidific sedimentation – fine and coarse deposits

Distance from river mouth (x)

Gra

in s

ize

(d)

Fin

e

C

oars

e

Rat

e of

dep

ositi

on (

v)

Rate of deposition (v)

Wat

er d

epth

(D

)

Grain size (d)

Coarse Fine

vd

vd

DT–A

wave baseAccumulationTransportation

Fig. 4.8 Illustration of major sedimentological and bottom-dynamics processes in lakes. (Modified from Håkanson & Jansson 1983.)

ESC04 9/5/06 3:03 PM Page 120


oxygen status of the sediments. Zoobenthos willnot generally survive if the oxygen concentrationat the sediment–water interface becomes lowerthan about 2 mg L−1. When the zoobenthos die,the biological mixing (bioturbation; see Fig. 4.10)of the sediments will stop. This has profoundconsequences for the lamination of the sedimentsand, hence, has implications for interpretationsabout the age distribution of sediments.

Generally, various types of zoobenthos live insediments down to about 5–15 cm sediment depth(see Fig. 4.11). This upper part of the sedimentcolumn is biologically active in the sense that the bottom fauna can influence the physical,chemical and biological conditions of the sedi-ments and cause bioturbation. Important groupsof bioturbators in lakes are large zoobenthossuch as worms and crustaceans. Benthic animalsare known to display great areal, temporal, ver-tical and species-specific patchiness within lakesand great variability among lakes. It has beenshown that bottom animals can eat sediment up to an average of seven times and this will,

evidently, strongly influence both the age andage distribution of the sediments (see Håkanson& Jansson 1983).

If the larger animals (macro- and meiofauna)die, bioturbation is halted and laminated (layeredand unmixed) sediments may appear. Continu-ous sedimentation will cause the sediment layerto grow upward so that the bioturbation limit,i.e. the limit between the upper biological layerand the lower biopassive (or geological) layer,moves upward. As stressed previously, bioturba-tion will influence the age and the age distribu-tion at any given sediment depth and this hasevident and profound implications for sedimentdating, i.e. when the sediments are used as a his-torical archive. When bioturbation is negligible,for example during anaerobic conditions, theage of the laminated sediments can be deter-mined in a very straightforward manner by‘counting varves’. Section 4.4 will discuss somemethods to determine the age of lake sedimentsand exemplify why this is important in lakestudies.

0 10 20 30 40 50 60 70 80 90 1000

10

20

30

40

50

60

70

80

90

100

Area (%)

Dep

th (

%)

Area A

DTA

Form factor (Vd )

A = f (Vd, Dmax, DTA, Area)

Vd = 3.Dm/Dmax

0.61.21.82.4

Area A = Area . ((Dmax – DTA)/(Dmax + DTA . exp(3 – Vd 1.5)))0.5/Vd

Fig. 4.9 An illustration of how the formfactor (= the volume development), Vd, can be used to express the form,here given by the relative hypsographiccurve (= the depth–area curve) of lakes.Shallow lakes with a small Vd have arelatively large area above the wave base (the surface-water areas), whereprocesses of wind/wave-inducedresuspension will influence the bottomdynamic conditions. Deep, U-formedlakes generally have smaller areas abovethe wave base (ET areas) and large deep-water areas. The equation describeshow the accumulation areas (Area A)may be calculated from data on lake area(Area), maximum depth (Dmax), thedepth of the wave base (the criticaldepth, DTA) and the form factor (Vd).(Modified from Håkanson 1999.)

ESC04 9/5/06 3:03 PM Page 121

122 LARS HÅKANSON

4.2.2.3.2 Compactional and diagenetic processes Anumber of important processes influence sedi-ment accumulation following deposition. Theseinclude diffusion, mineralization, compactionand burial. Diffusion is the chemical transport

of dissolved substances from sediment inter-stitial water back to lake water regulated byconcentration gradients. For many substances(e.g. phosphorus and caesium), the diffusivetransport is highly dependent on sediment

Sedimentation (cm/yr)

Sedimentdepth (cm)

x0

x1

x2

x3

x4

x5

x6

Water content;compactionW0

W1

W2

W3

W4

W5

W6

Bioturbation limit

Upwardbiotransport(Bu)

Downwardbiotransport(Bd)

Bu > Bd at allsediment depths

Substrate decomposition= mineralization

W1 > W2 > W3, etc.

Fig. 4.10 Illustration of key processes related to bioturbated sediments (sedimentation, upward and downward biotransport, substratedecomposition and compaction). (Modified from Håkanson & Jansson 1983.)

Anaerobic sediments

Aerobic sediments

Laminatedsediments(generally thinner, darker winter layers;thicker lightersummer layers)

Bioturbation(showing areal, vertical, temporal andspecies specific patchiness)

0

3

6

9(cm)

Bioturbationlimit(at 5–15 cm) Normal conditions Transition Polluted zone Very polluted

zone zone

Increased contamination of organic materialsDecreased oxygen concentration

Fig. 4.11 Bioturbation and laminatedsediments. Under aerobic (= oxic)conditions zoobenthos may create abiological mixing of sediments down toabout 15 cm sediment depth (thebioturbation limit). If the deposition oforganic materials increases and hencealso the oxygen consumption frombacterial degradation of organicmaterials, the oxygen concentrationmay reach the critical limit of 2 mg L−1,when zoobenthos die and bioturbationceases and laminated sedimentsappear. (Modified from Pearson &Rosenberg 1976).

ESC04 9/5/06 3:03 PM Page 122


redox-conditions – the lower the redox poten-tial (and oxygen concentrations), the higher thediffusive fluxes (see Håkanson 1999). Diffusionmay be a dominant flow of phosphorus in highlyproductive lakes. The function of bacteria inlakes is the decomposition of organic matter,and mineralization (which produces dissolvedsubstances during early diagenesis) is the namefor this process. Compaction in sediments con-cerns the vertical change in sediment water con-tent and bulk density due to the accumulatedweight from overlying sediments. The watercontent may change from about 85% in theuppermost sediment layer in a lake to about 70%at a sediment depth of 15 cm as a result of com-paction (see section 4.2.3 below).

As a result, substances deposited on a lakebed may be returned from sediments back towater by diffusive and advective (= resuspension)processes. There is, however, a sediment depthbeneath which substances will not return. Instead,they will be buried by the constant deposition ofnew matter.

4.2.3 Variations in lake sediment deposits

As lake sediments are influenced by materialssupplied from the catchment area (allochthonousmatter), from materials produced in the lake andfrom matter precipitated out of solution in thelake (autochthonous matter), there are great dif-ferences among lakes in sediment characteristics,such as elemental composition, physical propertiesand bottom fauna communities (see Håkanson

& Jansson 1983). This will be exemplified hereby presenting information in two tables.

Physical sediment characteristics in lakes ofdifferent trophic and humic status are shown inTable 4.6. The vertical changes in sediment watercontent (or organic content or bulk density)may be expressed by the following relationship:

W(x) = W0−1 + Ks · ln(2 · x)

where W(x) is the water content (in percentagewet weight – % ww) of a 1-cm-thick layer atsediment depth x (i.e. x ± 0.5 cm), W0−1 is thewater content (in % ww) of surficial sediment(0–1 cm) and Ks is an empirical sediment con-stant illustrating the change in water contentwith sediment depth.

If Ks attains a high numerical value (such as −10), it means that there is a strong vertical gradient, i.e. the water content decreases verysignificantly with sediment depth as a result of compaction. If Ks is small (such as −0.5), it is characteristic of lakes with loose sedimentsdeep down in sediment cores. One should notethat the Ks value is site-specific as well as lake-specific. The value depends on prevailing bottomdynamic conditions (erosion, transportationand accumulation). It varies comparatively littlewithin open water areas (areas of accumulation)in lakes, and much more in erosion and trans-portation areas. The value of Ks can vary widelybetween lakes depending on the chemical, phys-ical and biological characteristics of the lake.From Table 4.6 it can be noted that humic lakes

Table 4.6 Data from sediment cores from lakes illustrating the relationship between physical sediment character (the water content ofsurficial sediment, 0–1 cm, and the sediment constant illustrating the vertical gradient in sediment compaction) and lake type (as givenby the trophic and humic status) (From Håkanson & Jansson 1983.)

Lake Lake type Water content Sediment constant Sediment character

Ingen Polyhumic, oligotrophic 95.2 −0.41 Very loose, small vertical changesTrosken Polyhumic, oligotrophic 95.4 −0.83 Very loose, small vertical changesSkal Polyhumic, oligotrophic 97.4 −0.64 Very loose, small vertical changesHjalmaren Mesohumic, eutrophic 90.4 −2.99 Loose, clear vertical gradientFreden Mesohumic, eutrophic 87.7 −3.80 Loose, clear vertical gradientVasman Mesohumic, mesotrophic 95.9 −3.95 Very loose, clear vertical gradientAspen Mesohumic, mesotrophic 86.9 −5.78 Loose, strong vertical gradientVanern Mesohumic, oligotrophic 85.5 −5.97 Loose, strong vertical gradientVattern Oligohumic, oligotrophic 92.5 −13.6 Very loose, very strong vertical gradient

ESC04 9/5/06 3:03 PM Page 123

124 LARS HÅKANSON

generally have loose sediments with high watercontents and low Ks values, whereas eutrophiclakes often have sediments with a water contentof surficial sediments in the range 83–93% wwand Ks values of about −2 to −5. Mesotrophicand oligotrophic lakes cannot be distinguishedfrom eutrophic lakes by the water content of the sediments, but often by the sediment con-stant, which generally attains higher numericalvalues for low-productivity lakes. This depends,however, on many things, for example supply of minerogenic matter, which is relatively highin oligotrophic lakes, and on bioturbation, whichis generally higher (causing higher Ks values) in sediments rich (but not too rich) in organic matter and food for the bottom fauna.

Chemical analyses of sediment cores can offera good key to the history of lakes, and also to thepresent conditions. A chemical classification ofelements in lake sediments is given in Table 4.7.Sediments affect and reflect the characteristicsof lakes. Many studies have been presentedrelated to the chemical composition of lake waterand sediments and to interpretational codes of

what this means in terms of lake processes (see Håkanson & Jansson 1983; Pedley 1990).Table 4.8 is included here to stress this point. Itprovides data on sediment chemical properties(P, organic carbon, Fe, Mn, Ca, K, Si and Al) for lakes in Minnesota categorized into fourgroups (low and high organic, and low and highcarbonate groups), for four large African andfour large American lakes. From Table 4.8 it canbe noted that some of these substances do notvary a great deal among these lakes (e.g. P andSi), whereas other substances (such as Ca andorganic C) vary a great deal. This is an import-ant point. Sediment variables, just like watervariables, can vary among lakes owing to differ-ences in catchment area characteristics (geology,soils, land-use, vegetation, etc.) and within lakesdepending on climatological factors, season ofthe year, changing winds (such as frequency ofstorms) and bottom dynamic conditions. Thismeans that some variables can reflect typical lakeconditions much better than others, Sedimentphosphorus, manganese and iron are well known(see Table 4.5) for poorly reflecting typical lake

Table 4.7 A chemical classification of elements in lake sediments. (Modified after Kemp et al. 1976.)

1. Major elements (Si, Al, K, Na and Mg) make up the largest group of the sediment matrix2. Carbonate elements (Ca, Mg and CO3–C) constitute the second largest group; about 15% of the materials3. Nutrient elements (org.-C, N and P) account for about 10% in recent lake sediments4. Mobile elements (Mn, Fe, P and S) make up about 5% of the total sediment weight5. Trace elements (Hg, Cd, Pb, Zn, Cu, Cr, Ni, Ag, V, etc.), the smallest group accounting for less than 0.1% of the

sediments

Table 4.8 Chemical characteristics of lake sediments from different regions of the world (data from Jones & Bowser 1978).

Region Lake/group type P Organic C Fe Mn Ca K Si Al

Minnesota lakes Low organic 0.13 7.6 5.0 0.6 0.9 1.2 – –High organic 0.17 21 3.8 0.1 2.0 0.7 – –High carbonate 0.14 9.9 2.2 0.15 15.2 0.5 – –Low carbonate 0.08 6.1 2.5 0.4 6.8 0.4 – –

African lakes Kivu – – 5.3 0.09 9.5 – 19 5.0Tanganyika – – 5.0 0.03 1.2 – 26 10Edvard – – 2.4 0.03 3.0 – 29 4.3Albert – – 6.2 0.1 1.4 – 25 12

Great Lakes Ontario 0.07 – 3.7 0.06 0.4 2.3 24 5.1Erie 0.06 – 2.8 0.06 0.35 2.2 26 4.8Michigan 0.08 – 1.5 0.08 11 1.3 25 2.8Superior – 2.3 2.5 0.05 1.2 0.5 24 2.4

ESC04 9/5/06 3:03 PM Page 124


conditions, because the sediment concentrationsof these substances mainly reflect sediment redox-conditions and low redox values may appear in sediments of most lakes. On the other hand,sediment Ca concentrations more closely reflecttypical lake properties on a scale from low tohigh calcareous conditions.

4.2.4 Mass-balance modelling for lakes

The basic aim of mass-balance calculations is to quantify fluxes so that large and importantfluxes may be identified and differentiated fromsmall fluxes. The aim of this section, however, isnot to give a full mathematical account of all theprocesses shown in Fig. 4.2, only to give a briefintroduction to mass-balance modelling and toillustrate some basic concepts with a focus onsedimentation in lakes.

A simple mass-balance model for a lake isdepicted in Fig. 4.12. A typical mass-balancemodel envisions the lake as a ‘tank reactor’ inthe sense that the lake mixes completely duringan interval of time dt. The flow of suspendedparticulate matter (or a given contaminant) toand from such a lake, and net sedimentation maybe described by the following equation (this isthe steady-state solution to the equation given in Fig. 4.12):

0 = Q · Cin − Q · C − MW · Rsed · PF + MS · Rres(in) (out)(sedimentation)(resuspension)

where V is the lake volume (usually m3 or km3),C is the concentration of the substance in thelake water (units usually g L−1 or kg m−3; Cis the MW/V ), Cin is the concentration of the substance in the tributary (Cin has the samedimension as C), Q is the tributary water dis-charge to the lake (usually expressed as m3 yr−1

or m3 month−1), Rsed is the sedimentation rate of a given substance in the lake (like all rates,Rsed has the dimension 1/time and its unit is usu-ally 1/day, 1/month or 1/year), MW is the mass(= amount) of the substance in the lake water(units often in g or kg), PF is the particulate fraction (dimensionless), the only fraction thatcan settle out in lakes due to the influence ofgravitation, MS is the mass (= amount) of thesubstance on the lake bed (units in g or kg), Rresis the internal loading rate, or resuspension rate(units usually in 1/day, 1/month or 1/year). Thesteady-state assumption means that a change inlake concentration (dC) of the given substanceper unit of time (dt; usually in g L−1 month−1 orkg m−3 yr−1) is set to zero.

The lake water retention time (T, in days,months or years) is a fundamental concept inlake studies. It is defined as the ratio between the

MW

Lake water

MS

Inflow (Q . Cin) Outflow (Q . C = M . Rwat)

Water retention rate (Rwat)

Internal loadingrate (Rres)

Sedimentation rate (Rsed)

Particulate fraction (PF )

Sedimentation(MW . Rsed)

Internal loading(MS . Rres)

Lake sediments

V . dC/dt = Q . Cin − Q . C − MW . Rsed . PF + MS . Rres

dC/dt = change in concentration in lake waterC = MW/V; V = lake volumeTw = V/Q; Q = water dischargeCin = tributary concentration

Fig. 4.12 The basic mass-balance equation for a lakewith internal loading.

ESC04 9/5/06 3:03 PM Page 125

126 LARS HÅKANSON

lake volume and the water discharge to or fromthe lake (Q):

T = V/Q

The value of Q can be derived from time-seriesof measurements, or be given as a mean monthlyvalue, or as a mean yearly value. In the lattercase, T is generally referred to as the theoreticallake-water retention time.

The concentration in the lake water may thenbe expressed as:

C = (Q · Cin + MS · Rres)/(Q + V · Rres · PF)

This expression for the lake concentration isfundamental in lake science and management.When Vollenweider (1968) presented his firstloading model for lake eutrophication (phos-phorus), it meant a breakthrough not ‘just’ forlake management but also for lake modelling.Vollenweider simplified the mass-balance modelfirst by omitting seasonal variations and insteadgave the annual budget. He also omitted differentnutrients and different forms of the nutrients,and instead made the calculations for totalphosphorus. In addition, he disregarded inter-nal loading (the MS · Rres term in the given equa-tion) and simplified the sedimentation term (MS · Rsed · PF, which he approximated to √T).This gave the famous Vollenweider model:

C = Cin /(1 + √T )

It is evident that substances with large Rsedvalues settle rapidly, near the point of discharge,and that substances with small Rsed values maybe distributed over much larger areas. For mostsubstances, it is important to determine or pre-dict the Rsed value, which is related to the settlingvelocity, v (v = z · Rsed, where z is the distancethrough which the particle sinks in the giventime interval). The settling velocity, v, is gener-ally given in centimetres per second or metres peryear. Given Rsed or v, one can model or predictwhere high and low concentrations are likely toappear in water and sediments. This is the key topredicting where high and low ecological effects

may appear. It is important to note the differ-ence between the settling velocity (v in cm s−1)and the sedimentation rate (Rsed in L s−1).

Stokes’ law expresses the settling velocity (v) as:

v = ((dw – dp · g · d2)/(18 · μ · Ø)

where v is the settling velocity (usually in cm s−1 or m month−1), dp is the particle density(usually in g dw cm−3), dw is the density of thelake water (often set to 1 g ww cm−3), g is theacceleration due to gravity (980.6 cm s−1), d isthe particle diameter (in m, cm or mm), μ is thecoefficient of absolute viscosity (obtained fromstandard tables; 0.01 poise at 20°C) and Ø is the coefficient of form resistance (set to 1 forspheres; Hutchinson 1967).

Stokes’ law (Stokes 1851) is depicted in Fig. 4.13. The behaviour of material that followsStokes’ law (i.e. particles with a diameter betweenabout 0.01 and 0.0001 cm) differs from that ofthe coarser fraction material and from that of still finer material. The sedimentologicalbehaviour of the material is closely linked to thegrain size of the individual particles (Einstein1950; Allen 1970). The sedimentological beha-viour of the very fine materials is governed byBrownian motion. These latter particles are sosmall that they will not settle individually, butwill do so if they form larger flocs or aggregatesthat are dense enough to settle according toStokes’ law (Kranck 1973, 1979; Lick et al. 1992).The cohesive materials that follow Stokes’ law arevery important, because they have a great affinityfor pollutants (Fig. 4.7). This group includesmany types of detritus, humic substances andplankton. All play significant roles in aquaticecosystems (Salomons & Förstner 1984).

The settling velocity (v) of a given particle,aggregate or particulate pollutant, and its dis-tribution in a lake, depends on the density, sizeand form of the particle (and on the hydro-dynamics of the flow of water in the lake). If the particle density, dp, is close to 1, if the formfactor, Ø, is large and if the diameter, d, is small, the settling velocity, v, and the sedimenta-tion rate, Rsed, may be very slow. If Rsed is closeto 0, the particle or aggregate is conservative in

ESC04 9/5/06 3:03 PM Page 126


the sense that it may not be deposited in the lake.In many lakes (see Håkanson & Jansson 1983),Cl, Ca and alkalinity are typically conservativesubstances, and colour, organic matter, total P,Si, particulate P and suspended matter are morereactive.

4.2.4.1 The distribution coefficient

A very important part of most mass-balancemodels for chemical substances is the distribu-tion coefficient. The lake distribution coefficientis also often called the partition or partitioningcoefficient and gives the particulate fraction (PF)or the dissolved fraction (DF = 1 − PF). Relatedto sedimentological processes, there are threemajor categories of matter in lake water:1 the particulate fraction (PF), which is the onlyfraction settling by gravitation;2 the dissolved fraction, which is the most im-portant one for direct biouptake by, for example,phyto- or zooplankton and often referred to asthe bioavailable fraction;3 the colloidal fraction, which is neither subjectto direct uptake nor to sedimentation becausethe particles are too small.

The particulate fraction and the colloidal fraction may be subject to mineralization bybacterioplankton. Traditionally (see Santschi & Honeyman 1991; Gustafsson & Gschwend1997), the Kd concept is used in these contexts;Kd is the ratio between the particulate (C ′par in gkg−1 dw) and the dissolved (Cdiss in g L−1) phases,i.e. Kd = C ′par /Cdiss. The Kd ratio is often given in L kg−1. This means that the dissolved fraction(Ddiss) can be written as:

Ddiss = 1/(1 + Kd · SPM · 10−6)

where SPM is the amount of suspended matterin the lake water (in mg L−1). It is essential todistinguish between the dissolved and the par-ticulate fractions for all substances. It is espe-cially important to do so for the key nutrient inlake management, phosphorus, because phyto-plankton takes up dissolved P and only particulateP can settle out. This means that there are differ-ent transport routes for the two fractions. Theconcentration of suspended particulate matter(SPM) influences the distribution of phosphorusinto these two fractions. The settling velocity for particulate P in (m yr−1) may be turned into a

100

10

1

0.1

0.01

0.001

0.0001

0.00001

0.0000010.0001 0.001 0.01 0.1

m month−1

26,000

2,600

260

26

2.6

0.26

Particle diameter

Set

tling

vel

ocity

(cm

s−1

)

Cohesive materialsSuspended materials(Brownian motion)

Friction materials

1.1

1.52.0

2.5

3.0

Particle density (g cm−3)

Upper limit ofStokes's Law

Turbulent flow

Laminar flow

Fig. 4.13 The relationshipbetween the settling velocity(v; of spherical particles) inwater, particle diameter andparticle density (at 20°C) asgiven by Stokes’ law.

ESC04 9/5/06 3:03 PM Page 127

128 LARS HÅKANSON

sedimentation rate (dimension 1/time) by divisionwith the mean depth of the lake. The sedimenta-tion rate regulates sedimentation, and hence alsointernal loading in lakes. Figure 4.14 illustratesthe very important role that the PF value playsfor the concentration of phosphorus in water andfor sedimentation of phosphorus. It is evidentthat the PF value influences the concentration inlake water: the higher the particulate fractionand the less in dissolved form, the higher thesedimentation and the lower the concentration inwater, and vice versa. This is a general principlevalid for all substances.

4.3 PROCESSES AND IMPACTS OF NATURAL AND

ANTHROPOGENIC DISTURBANCE EVENTS

4.3.1 Storms and mass movements andearthquake records in lake sediments

Many studies have reported how storms, turbid-ity currents, earthquakes, landslides, etc. influencelakes and cause clearly distinguishable layers in the

sediment record (Sturm 1975; Sturm & Matter1978; Ludlam 1981; Eden & Page 1998; Ambers2001). These ‘turbidites’ can be deposited at anyseason of the year and may often be separatedfrom ‘ordinary’ lamina by a greater thicknessand often by a different colour or grain size distribution. It should be noted that turbiditycurrents triggered by subaquatic landslides cantravel very fast and influence large parts of lakes(Inouchi et al. 1996; Evans & Slaymaker 2004).They generally appear more frequently in largeand deep lakes; large and shallow lakes aredominated by wind-induced resuspension.

4.3.2 Lake-level fluctuations

Water-level fluctuations are common in manyreservoirs, but also in natural lakes dependingon changes in, for example, land-use practicesor precipitation (Thompson & Baedke 1995;Kato et al. 2003). The situation in Lake Aral is a famous example of this, but there are manyless well-known situations, for example LakeKinneret, Israel (Håkanson et al. 2000). Evid-ently, there are close links between changes inwater level and lake morphometry. If the waterlevel goes down, the lake area will be smaller,the wave base lower and the sediments and substances continuously accumulated below the previous wave base will be exposed andinfluenced by wind-induced wave activities. So,resuspension is likely to increase for some timeafter the lowering of the wave base. Connectedto this, the concentration of suspended parti-culate matter (SPM) is also likely to increase.This means that the water clarity will also bereduced, and this will influence the primary pro-duction of phytoplankton and benthic algae. Ifthere are changes in primary production thereare also probable alterations in secondary pro-duction of zooplankton and fish.

4.3.3 Lake sediment pollution

This section addresses the transport of pollu-tants to, within and from lakes. In soils, water isthe main transport medium for pollutants andthe migration of chemical pollutants is therefore

Con

cent

ratio

n in

wat

er (

C in

µg

L−1) 30

15

0

60

30

0

Sed

imen

tatio

n (S

ed in

kg

yr−1

)

0 0.50 1.00100%dissolvedphase

100%particulatephase

Particulate fraction (PF )

C

Sed

Fig. 4.14 Calculations illustrating the important role of the particulate fraction in mass-balance calculations (annual simulations for a lake with an area = 1 km2; mean depth= 10 m; catchment area = 10 km2; mean annual precipitation =650 mm yr−1; mean tributary concentration = 26 μg L−1 of totalphosphorus; and a settling velocity of 5 m yr−1 for the particulatefraction).

ESC04 9/5/06 3:03 PM Page 128


tightly coupled to the hydrocycle. This sectionattempts to give a general introduction of the basicprinciples and processes regulating the migrationof chemical pollutants. Possible anthropogenicsources of pollutants and types of deposition to the ground will be mentioned only briefly.The sources generally can be divided into activ-ities related to: (i) combustion (e.g. for energyproduction or waste incineration), (ii) technicalprocesses (e.g. metal and chemical industries)and (iii) other anthropogenic additions (e.g. distribution of pesticides and fertilizers) (e.g.Baudo et al. 1990; Butcher et al. 1992).

Different pollutants have different ranges of atmospheric distribution, owing to variableatmospheric residence times (e.g. mercury has a relatively long residence time in the atmo-sphere compared with many other metals). Thespatial scales may be local, regional or global.After transport in the atmosphere pollutants are deposited either together with rain (wetdeposition) or without (dry deposition). The rateof atmospheric deposition depends on the typeof pollutant and soil cover. A coniferous forest,for example, has a large active surface area persquare metre forest floor and generally catchespollutants effectively. Pollutants are transferredto the soil as direct deposition, throughfall or litterfall. After deposition to the ground, thepollutants mainly migrate through the environ-ment with water. It is, therefore, of interest tostudy processes influencing the transport fromthe watershed divide to the lake.

Some elements (such as copper) are micro-nutrients and essential for life. Such elementsare needed within a given concentration rangeand are toxic outside the range. The effect is,thus, a function of concentration or load. Differ-ent ecosystems do not, however, necessarilyshow the same response to the same given load.Sensitivity (or vulnerability) parameters includewater pH, water hardness or lake colour, whichwill influence the effect of a given pollutant con-centration (see section 4.6.2). Many pollutantssuch as cadmium, mercury and most synthet-ically manufactured organic pollutants do nothave any known physiological functions. Theyare non-essential to life.

During migration through ecosystems, pol-lutants might be physically or chemically trans-formed and more or less toxic forms may appear.There are three main pathways of transformation.1 Degradation including (i) photolysis, mediatedby sunlight and (ii) biologically mediated mineral-ization of organic matter into its inorganic basiccomponents (energy locked within more com-plex chemical structures is used as a source ofenergy by degrading microorganisms).2 Radioactive decay of radionuclides at the rateof their physical half-lifes. The decay producesdaughter nuclides, which in turn decay. Thiscontinues until a stable element is reached.3 Chemical speciation (e.g. oxidation/reduction,metal methylation) due to changed environ-mental conditions.

For mass transport of chemical pollutants thereare two basic physical processes, diffusion andadvection. Nature has an inherent reluctanceagainst structured orders, such as differences inconcentrations, for example across different inter-faces such as air–water and water–sediment.This is one interpretation of the second law ofthermodynamics and the force of diffusion. Thediffusive flux of pollutants is, thus, proportionalto the difference (gradient) of concentrations.Volatilization/evaporation is one example of a diffusive flux. Advection, on the other hand, is the mixing by stirring. The force might beexternal, for example wind-driven waves in lakes,or internal, for example density differences instratified waters, which causes mixing. Advectivetransport can be laminar or turbulent.

Retention is the opposite to migration, a termexpressing the storage of a pollutant within thesystem in relation to the load. Pollutants withlong retention times (a high retention) are poten-tially more hazardous because it means longerexposure times within the ecosystem. The reten-tion is affected by several factors. Some of themost important are the chemical characteristicsdetermining the distribution coefficient, Kd, andthe physical factors influencing the flushing ofpollutants through the system. Influences of thephysical environment on the retention of pollut-ants mainly include the rate of flushing thoughthe system. This is expressed by the theoretical

ESC04 9/5/06 3:03 PM Page 129

130 LARS HÅKANSON

water residence time T. The theoretical residencetime of a substance, Tr, is defined by:

Tr = (V/Q)/(C/Cin)

where C is the concentration in the system andCin is the concentration in the inflow.

The residence times are related to time-dependent processes such as the settling of fine-grained particles trough the water column andthe sorption of pollutants to carrier particles(and other chemical processes), which are ex-amples of processes favoured by long residencetimes.

4.3.4 Toxicity of chemical water and lakesediment pollutants

In many contexts of lake management, there is a focus on the ecosystem-scale (i.e. on entirelakes), and on the following three major chem-ical threats to aquatic ecosystems (Table 4.9): (i) acidification, (ii) eutrophication and (iii) contamination (by metals, organic toxins andradionuclides). This section will examine funda-mental principles and processes regulating thespread, biouptake and ecosystem effects of con-taminants (see Munawar & Dave 1996).

The well-known environmental pollutantmercury belongs to a group of elements oftenreferred to as heavy metals (i.e. metals with adensity > 5 cm−3). These metals generally formoxides and sulphides, which are often very hardto dissolve, and they tend to be bound in stablecomplexes with organic and inorganic particles,

the ‘carrier particles’, which is an importantconcept in sediment–water systems. The greatinterest in heavy metals in aquatic ecotoxico-logy derives from the fact that some of these elements are supplied to water systems in greatexcess by humans, and that some of them arehazardous to aquatic life (see Bowen 1966;Förstner & Müller 1974; Förstner & Wittmann1979; Salomons & Förstner 1984).

A traditional way of determining toxicity ofmetals and other toxins is to establish the LC50or LD50 value, where LC stands for lethal con-centration and LD for lethal dose. The value isobtained for the concentration that exterminates50% of the test sample relative to a control groupof test organisms during a certain time span.More than 200 monographs on various toxico-logical test systems have been published (e.g.Cairns 1981; Burton 1992). A crude rule of thumbstates that the least hazardous elements appearwith the highest concentrations in water, sedi-ments and biota, and vice versa, the ‘abundanceprinciple’ (see Håkanson 1980). Elements appear-ing on the ppb-scale (parts per billion, 109), i.e.with extremely low natural concentrations, are,for example Hg, Ag and Cd (Table 4.10). Ele-ments on the ppm-scale (106) are, for example As,Co, Cr, Cu, Mo, Ni, Pb, Sn, V and Zn. Elementson the mg-scale (103) are, for example Al, Ca,Fe, K, Mn and Na. Pollutants are also classifiedaccordingly into water soluble (hydrophilic ele-ments and compounds) and organic (soluble inorganic solutes; liphophilic elements and com-pounds). Liphophilic compounds are generally‘bioavailable’.

Chemical threat Ecological effects

Acidification Increase in filamentous algaeReduced reproduction of crustaceans, snails, bivalves and roach

Eutrophication Decrease in Secchi depthIncrease in chlorophyll a and hypolimnetic oxygen demand

Contamination:metals Increased concentration in fish for human consumptionradionuclides Decrease in reproduction of key organisms, e.g.

zooplankton, benthos and fishorganic toxins

Table 4.9 Different chemical threatsto aquatic ecosystems and someexamples of ecological effect variables.There are also many physical threats toaquatic ecosystems, like the building ofdams, piers and marinas, and manybiological threats, such as theintroduction of new species.

ESC04 9/5/06 3:03 PM Page 130


Surfaces of fine-grained particles are chemic-ally active. Surface sites are positively or nega-tively electrically charged or neutral. Adsorptionreactions between charged surfaces and pollut-ants in ionic form include different types ofphysical/chemical bindings. Pollutants may occuras free uncomplexed species or as various com-plexes. Pollutants are attracted to different typesof carrier particles depending on their inherentchemical properties. Fine-grained carrier particleshave a large surface area per weight unit and are,thus, important for the migration of pollutants.Carrier particles exist in a wide size spectrum.From 1 nm size (10−9 m) up to approximately0.45 μm (10−6 m) the particles are generallycalled colloids. From 0.45 μm to about 0.15 mm,particles suspended in water settle according toStokes’ law by laminar flow (cohesive material).It should also be noted that suspended particles areconstantly moving and they aggregate (flocculate)and disaggregate as a result of collisions and reac-tions with other particles and fluid shear. Thecarrier particles may be classified into organic(e.g. humic matter) and inorganic (e.g. clays,(hydr-)oxides, Fe/Mn precipitates) or accordingto genetic origin (e.g. biogenic (mainly organic)or pedogenic (mainly inorganic matter)).

The chemical environment can be expressedin terms of, for example, pH (i.e. the availabil-ity of H+ ions), oxygen (O2) and redox potential

(Eh). These factors influence the pollutants in a number of ways. The formation of import-ant carrier particles, for example iron (Fe) and manganese (Mn) precipitates, are, for example,enhanced by high pH and Eh. The presence ofO2 determines the presence of oxides, such asFe-oxides, or reduced species, which is import-ant for pollutant cycling in lakes. The survival of higher life forms, such as zoobenthos, alsodepends on the supply of oxygen. The pH valuehas a major influence on many chemical/physicalprocesses, some of which are important for themigration of pollutants. The H+ ions competefor carrier particle binding sites with pollutants.A lower pH generally decreases the Kd value.Flocculation of carrier particles also depends on pH. With increased pH the usually negativecharges of natural particles increase and the particles become less likely to flocculate.

The distribution form of the metal in aquaticenvironments is very important for the toxicityand the potential ecosystem effects (Gottofrey1990; Wicklund 1990). Generally, the toxicityis highest for the ionic species and propor-tional to the oxidation number, for exampleCrO4

− is more toxic than Cr3+. The potentialtoxic effects of metals can often be signific-antly reduced because the metals are bound todifferent compounds, which may camouflagethe toxic properties. Low pH and high redox

Element Igneous rock Soils Fresh water Land plants Land animals

Ag 0.07 0.1 0.00015 0.06 0.006Al 80,000 70,000 0.25 500 4–100As 1.8 6 0.004 0.2 ≤ 0.2Cd 0.2 0.06 < 0.08 0.6 ≤ 0.5Co 25 8 0.0009 0.5 0.03Cr 100 100 0.0002 0.25 0.075Cu 50 20 0.01 15 2.5Fe 55,000 40,000 0.65 140 160Hg 0.08 0.03–0.8 0.00008 0.015 0.045Mn 1000 900 0.01 650 0.2Mo 1.5 2 0.00035 0.9 < 0.2Ni 75 40 0.01 3 0.8Pb 15 10 0.005 2.5 2Sn 2 10 0.00004 < 0.3 < 0.15V 150 100 0.001 1.5 0.15Zn 70 50 0.01 100 150

Table 4.10 The abundance of variouselements (in ppm) in igneous rocks,soils, fresh water, land plants and landanimals (from Bowen 1966).

ESC04 9/5/06 3:03 PM Page 131

132 LARS HÅKANSON

potential (Eh) often increase metal toxicity. The solubility of most heavy metals increases atdecreasing pH and metals that previously havebeen bound in rather harmless particulate formsin the sediments may be recirculated to the lake water and express their toxic properties ifsediment pH or Eh changes. The roles of heavymetals are, as stressed, complicated by the factthat some of them are essential in small amountsfor the organisms.

Table 4.11 gives one example of how to struc-ture the very complex group of organic toxins.For example, AOX stands for adsorbed organic-ally bound halogen, TOCl for total chlorinatedorganic material, EOCl for extractable organ-ically bound chlorine, EPOCl for extractable persistent organically bound chlorine, etc. Thefocus of many studies has been on emissions of chlorinated substances from paper and pulpmills (see Södergren 1992). In sediments, onlyabout 2% of a sum-parameter such as TOClconsists of EOCl, and only a small fraction (< 1%)of EOCl consists of chemically identified sub-stances, specific parameters, such as dioxins,polychlorinated biphenyls (PCBs) or dichloro-diphenyltrichloroethane (DDT). The toxicity oforganic pollutants, such as EOCl, DDT andPCB, may be manifested in many different ways,for example increased fin erosion in perch,increased frequencies of skin ulcers in herringand increased skeleton deformations, such asdeformed jaws in pike or spinal column bendingin fourhorn sculpin (Bengtsson 1991). Södergren

(1992) gives a thorough evaluation of biologicaleffects of bleached pulp mill effluents in theBaltic.

In summary, the major threats from chemicalpollutants today concern nutrients (phosphorusand nitrogen) causing different types ofeutrophication effects (Ambio 1990; Wallin et al. 1992), toxic substances such as metals(Cd, Pb, Cu; Förstner & Müller 1974), chlori-nated organics (e.g. PCBs, DDT, dioxins;Södergren et al. 1988) and acidification of landand water, its ecological damage and economicconsequences (Ambio 1976; Likens et al. 1979;Merilehto et al. 1988). Case Study 4.1 gives anexample of EOCl contamination.

4.3.5 Climatic change

Many lakes are highly sensitive to changes intemperature and precipitation (both magnitudesand frequencies; see Intergovernmental Panel onClimate Change (IPCC) website), because this mayinfluence fundamental processes related to lakeecosystem structure and functioning (e.g. ratesof evaporation, lake water level and productionand biomass of key functional groups or species).Under extreme climatological conditions, veryshallow lakes may even disappear. Responses to climate change will also vary between lakes at different latitudes, altitudes and dependingon physical geographical conditions. Lakes, andespecially lake sediments, are good sources ofinformation about past climatic/environmental

Table 4.11 Some well-known chlorinated organics.

Contaminant Details

TOCl Total organically bound chlorineAOX Adsorbed organically bound halogenEOCl Extractable organically bound chlorineEPOCl Extractable (acid-)persistent organically bound chlorineDioxins Polychlorinated dibenzo dioxins (PCDD); and furans; there are many dioxins and furans, of which,

‘the dirty dozen’, are considered of special interest in ecotoxicologyPCB Polychlorinated biphenyls; lipophilic substances used, e.g. in oils; certain forms, such as planar PCBs

are considered to be responsible for the sterility of Baltic sealsDDT Dichlorodiphenyltrichloroethane; this group includes, e.g. lindane, aldrine, dieldrine and

dichlorodiphenyldichloroethylene (DDE), all well-known from R. Carson’s book Silent SpringHCB HexachlorobenzeneHCH Hexachlorocyclohexane

ESC04 9/5/06 3:03 PM Page 132


Case study 4.1 Organic toxins in sediments: Baltic Sea

Discharges of chlorinated organic materials from paper and pulp mills (PPMs) have attracted agreat deal of attention in Sweden and many parts of the world (Södergren 1992). This sectionwill give a sedimentological angle to this important environmental problem. Case Fig. 4.1a illus-trates the geographical distribution pattern of EOCl in surficial (0–1 cm) sediment samples fromaccumulation areas, and Case Fig. 4.1b shows how extractable organically bound chlorines(EOCl) in sediment samples co-vary with two very important and specific toxins – dioxins andfurans. Other features are evident from Case Fig. 4.1a.1 There is a large-scale spread of EOCl in this water system, the Baltic. The larger the emissions,the larger the impact areas and the greater the potential ecological problems. The distributionof EOCl in surficial sediments shows a very characteristic pattern, with high concentrationsclose to the PPMs along the coast.

(c)

Area of diffuseinfluence

Coastal jet-zone

Coastal area withhigh contamination

(a)

(d)

2000

1800

1600

1400

1200

1000

800

600

400

200

00 5 10 15 20 25 30 35

Distance from Iggesund (km)

DioxinFuran

7000

6000

5000

4000

3000

2000

1000

0

EOCl

Dioxins (pg g−1 org)Furans (pg g−1 org)EOCl (µg g−1 org)

(b)

0

−2

−4

−6

−8

−10

−12

−14

−16

−18

−20

Sed

imen

t dep

th (

cm)

0 2 4 6 8 10 12 14 16 18 20EOCl (µg g−1 dw)

Baltic proper

1985

1970

~1950

~1930

8 cores and 38 surface samples1000

500

250

100

50

0

EOClin surficial sedimentsfrom A areas in µg g−1 LOI

Case Fig. 4.1 (a) Distribution of EOCl in surficial A-area sediments in the Baltic. (b) The relationship between EOCl, dioxinsand furanes in surficial A-area sediments taken at different distances from the Iggesund paper and pulp mill in the Bothnian Sea.(c) Illustration of the coastal jet-zone and the major hydrological flow pattern in the Bothnian Bay and the Bothnian Sea. (d) Thehistorical development of EOCl in the Baltic (based on eight sediment cores and 38 surficial sediment samples). (From Jonsson1992; Håkanson 1999.)

ESC04 9/5/06 3:03 PM Page 133

134 LARS HÅKANSON

conditions. Case Study 4.2 sets out an exampleof climate change impacts upon Lake Batorino,Belarus.

4.4 SEDIMENT DATING AND SEDIMENT RECORDS

4.4.1 Sediment dating using radionuclides

The sediments reflect what is happening in thewater mass and on the bottom – they may be

regarded as a tape-recorder of the lake’s histor-ical development and are often called ‘the geo-logical archive’ (Zolischka 1998; Hammarlundet al. 2003). The sediments also affect the condi-tions in the water via, for example, resuspensionprocesses and by the fact that the animals livingin the sediments play a fundamental role inthe ecosystem (Fig. 4.5). By extracting sedimentcores and conducting a number of analyses,information is obtained on changes that havetaken place in the ecosystem (Thomas et al. 1976;

2 The dominating water circulation in each basin (the Bothnian Bay, the Bothnian Sea and the Baltic Proper) constitutes an anticlockwise cell, which distributes the settling particles, thesuspended material and the pollutants in a typical pattern, reflecting the flow of the water (CaseFig. 4.1c). This anticlockwise cell is created by the rotation of the Earth (the Coriolis force), whichdeflects any plume of flowing water to the right in relation to the direction of the flow in theNorthern Hemisphere (and to the left in the Southern Hemisphere). Thus, when a Swedish riverenters the Baltic, the water turns to the right and follows the shore. The net hydrological flow is tothe south on the west (Swedish) side of the Baltic. The currents are rather strong and stable closeto land and weaker towards the centre of each basin. The figure illustrates only the net componentof the flow – this means that the water would also flow in most other directions during the year.3 A large number of sediment cores have been taken from Baltic accumulation areas and analysedfor EOCl. Case Fig. 4.1d gives results reflecting average conditions. The increase in the EOClconcentration in the sediments, and hence the increase in sediment and water contamination,started in the late 1950s and coincides well with the general contamination and increase ineutrophication in the Baltic.

It should be noted that most of the emissions of EOCl to the Baltic have now been halted as a result of legislation and public awareness of the problem. The Baltic is recovering from thiscontamination. The time perspective of this recovery can be illustrated using the sediment datain Case Fig. 4.1d. It will probably take 20–40 years until the system has recovered so that theEOCl concentration in surficial sediments is back to the ‘normal’ values that characterized thesystem about 50 years ago.

This example illustrates the very close and important connection between pollution site andload, distribution of pollutants by water currents, which regulate where high and low contamina-tion appears, and hence also where small and large potential ecosystem effects may be expected.This also shows that different contaminants from the same pollution site often (as in this case)are distributed in the aquatic environment together, because the distribution depends on thefact that both the dissolved and the particulate forms of the pollutants are distributed by thesame hydrodynamic and bottom-dynamics processes.

Relevant reading

Södergren A. (Ed.) (1992) Bleached Pulp Mill Effluents. Composition, Fate and Effects in the Baltic Sea. Report4047, Swedish Environmental Protection Agency, Stockholm, 150 pp.

ESC04 9/5/06 3:03 PM Page 134


Case study 4.2 Impact of global change on lake characteristics: Lake Batorino, Belarus

The presuppositions for this ‘global change’ scenario have been given by Håkanson et al. (2003).Wick (2000) has discussed how global temperature changes can influence vegetation and sedimentrecords. The basic assumption is that global warming would increase the mean annual temperature(as given by the curve in Case Fig. 4.2a). Note that this is a simulation of a hypothetical suddenincrease in temperature and that, realistically, global warming will lead to more gradual changes.In this scenario, the focus is on the final results rather than the path to the final results. Also notethat global temperature changes may cause more extreme seasonal temperature variations. In thisscenario, the mean annual temperature is raised by 2°C and an increased seasonal temperaturevariation has been assumed by applying an exponent > 1 for the weekly epilimnetic temperatures

30(°C)

Epi

Tem

p

15

Driving variable

0

(%)30 (c) (d)

Alarm limit

(a) (b)

Mac

roph

yte

cove

r

Pre

dato

ry fi

sh b

iom

ass

27

24

100,000

(kg ww)

(e)

50,000

0

1 261

Macrophyte cover

Secchi depth

521Weeks

781 1041

1 261 521Weeks

781 1041

Bac

terio

plan

kton

bio

mas

s

300,000

(kg ww)

(f)

150,000

01 261 521

Weeks781 1041

1 261 521Weeks

781 10410

2

4

0

5 Critical limit

Sec

chi d

epth

Alg

al v

olum

e

10(m) (mm3 L−1)

1 261 521Weeks

781 1041

140(µg L−1)

Tota

l pho

spho

rus

70

‘Global change’; Lake Batorino, Belarus

01 261 521

Weeks781 1041

Case Fig. 4.2 Case study of the effects of global temperature changes on Lake Batorino, Belarus, assuming that there would bea hypothetical change in lake water temperatures (a). How would this influence fundamental lake characteristics, such as lakeconcentrations of phosphorus (b), macrophyte cover and water clarity (Secchi depth) (c), algal volume in relation to operational‘critical’ and ‘alarm’ guideline values (d), biomass of predatory fish (e) and biomass of bacterioplankton (f)?

ESC04 9/5/06 3:03 PM Page 135

136 LARS HÅKANSON

(i.e. EpiTexp). In the following scenario, this exponent is set to 1.1. The temperature for the first10-year period (weeks 1–521) has been kept normal and then for the second 10-year period awell-tested lake foodweb model (LakeWeb, see Håkanson & Boulion 2002) has been used topredict the response of the system, here Lake Batorino, Belarus (see Case Table 4.2 for lake data).

Under these presuppositions, Case Fig. 4.2a gives the driving variable, lake epilimnetic temperatures. The question is, how will this change in temperature regime influence import-ant variables for lake management, such as lake total phosphorus (TP) concentrations (CaseFig. 4.2b), macrophyte cover and Secchi depth (Case Fig. 4.2c), algal volume (Case Fig. 4.2d),total fish biomass (Case Fig. 4.2e) and bacterioplankton biomass (Case Fig. 4.2f)? The follow-ing can be noted from Case Fig. 4.2.• An increase in lake temperature will increase lake TP concentrations significantly, and forseveral reasons. More phosphorus, for example, will be bound in organisms with short turn-over times (phytoplankton, bacterioplankton, benthic algae and zooplankton), which areincluded in the value given for the total lake TP concentrations. As a result an increase in meanannual temperatures will probably produce eutrophication (Case Fig. 4.2b).• This also means that warmer climatological conditions will reduce the Secchi depth andhence also decrease the macrophyte cover (Case Fig. 4.2c). The clearer the water, the higher themacrophyte production.• Warmer conditions will imply that algal volumes are likely to increase to values higher than the‘critical’ limit (set by national environmental agencies) during the growing season (Case Fig. 4.2d).• Warmer conditions will increase the fish production. This means a significant change in thestructure of the lake foodweb (Case Fig. 4.2e).• There are also clear changes for bacterioplankton biomass: the warmer the water, the morebacterioplankton will be produced (Case Fig. 4.2f).

The ‘global change’ scenario for Lake Batorino, therefore, indicates that major changes in thelake foodweb can be expected if the climate becomes warmer. A global warming in northernlakes would probably produce conditions similar to eutrophication. It would mean higher primary and secondary production, lower Secchi depths and hence also a reduced macrophytecover. This may seem self-evident, but it is not evident how this is manifested in terms of quantitative changes in key functional groups of organisms and in target variables for lakemanagement. Such changes, can, however, be calculated with the LakeWeb model.

Relevant reading

Håkanson, L. & Boulion, V. (2002) The Lake Foodweb – Modelling Predation and Abiotic/Biotic Interactions.Backhuys Publishers, Leiden, 344 pp.

Håkanson, L., Ostapenia, A., Parparov, A., et al. (2003) Management criteria for lake ecosystems applied to case studies of changes in nutrient loading and climate change. Lakes and Reservoirs: Research andManagement 8, 141–55.

Wick, L. (2000) Vegetational response to climatic changes recorded in Swiss late glacial lake sediments.Palaeogeography, Palaeoclimatology, Palaeoecology 159, 231–50.

Case Table 4.2 Data on the case-study lakes. (From Håkanson & Boulion 2002.)

Lake Catchment Area Mean Maximum pH Colour Cin Latitude Altitude Precipitation(km2) (km2) depth depth (mg Pt L−1) (μg P L−1) (°N) (m a.s.l.) (mm yr−1)

(Dm, m) (Dmax, m)

Batorino 92.5 6.3 3.0 5.5 8.0 54 120 54.5 165 650Miastro 133 13.1 5.4 11.3 8.0 31 80 54.5 165 650

ESC04 9/5/06 3:03 PM Page 136


Golterman et al. 1983; Jonsson 1992). This section gives a brief discussion on methods todetermine ongoing sedimentation and the age of recent sedimentary deposits and will not dis-cuss palaeolimnological methods (such as pollenanalysis and the radiocarbon method; see Reeves1968; Goudie 1981).

There are two very useful and widely appliedmethods to determine the age of recent lake sediments, lead-210 (210Pb) analysis (Robbins1978; Appleby & Oldfleld 1979; Legesse et al.2002) or the analysis of radiocaesium (137Cs; see Pennington et al. 1973; Yan et al. 2002).Lead-210 has a physical half-life of 22.3 yearsand 137Cs a half-life of 30.2 years, which makesthese substances very useful for dating recentsedimentary deposits. To determine the sedi-ment age using the 210Pb method, one must alsoquantify all fluxes from the catchment to thelake, internal fluxes of 210Pb (sedimentation andresuspension) and lake outflow. This means thatmass-balance calculations are essential. Sedi-ment dating using 137Cs is simpler. For this, it isimportant to have access to a fallout curve, suchas the one shown in Fig. 4.15 for a Finnish site.Radiocaesium in western Europe has two basicsources, fallout from the nuclear weapons test-ing (mainly between 1957 and 1975) and theChernobyl accident (April–May 1986). By tak-ing sediment cores and trying to identify thesepeaks in radiocaesium activity/concentration, agood estimate of the mean sedimentation rate and

the mean sediment age can be obtained. Evid-ently, this method is more likely to give moreaccurate estimates of sediment age in laminatedthan in bioturbated sediments.

4.4.2 Sediment dating using sediment traps

Sedimentation and sediment age can also bedetermined using sediment traps (see Håkanson& Jansson 1983). Generally, sediment traps are cylinders with a width:height ratio of 1:3placed vertically in the water. Figure 4.16 givesan example where sediment traps have been usedfor sediment dating. From this figure, it is evid-ent that the contamination of a river pollutant(here mercury entering Lake Ekoln, Sweden, fromRiver Fyris) and the content of the contaminantin lake sediments offer an excellent key to thepollution history of the lake. In the same way,the areal distribution pattern of the pollutant in lake sediments can be used to evaluate trans-port patterns in the system and to identify thepolluting site (or river).


4.5.1 Environmental consequence analysis andecosystem indices

In environmental management it is importantnot to use personal viewpoints as criteria to rank

1800

1600

1400

1200

1000

800

600

400

200

0

Fallo

ut o

f 137 C

s (B

q m

−2)

1 25 49 73 97 121 145 169 193 217 241 265 289 313 337 3611957 1959 1961 1963 1965 1969 1971 1973 1975 1977 1979 1981 1983 1983 1985 1987

Chernobyl fallout

River Kemijokki, Finland

Nuclear weapons test falloutFig. 4.15 Radiocaesium falloutfrom the nuclear weapons test andthe Chernobyl accident on thecatchment area of River Kemijokki,Finland (data from Ritva Saxen).These data can be used for datinglake sediments in this part of theworld. Note that in bioturbatedsediments, the age distribution atvarious sediment depths in asediment core would be levelled out,whereas in laminated sediments thefallout peaks may be more easilyidentified and the dating more exact.

ESC04 9/5/06 3:03 PM Page 137

138 LARS HÅKANSON

threats as a basis for action, but to have a moreobjective approach. There is a growing aware-ness that much better individual ‘indicators’ and aggregated ‘indices’ of environmental health are necessary because they alone can provide a rational structure for decision-making in the environmental sciences (Bromberg 1990;OECD 1991). An index (an aggregated measure)is generally distinguished from an indicator (a single variable), and an ecosystem (a singleinstance, such as a lake or a field) from an eco-system type (the summation of several to manyecosystems).

A frame of reference is required to assess the status of the environment. Since 1987, manycountries have accepted ‘sustainable develop-ment’ as a goal for environmental and economicpolicy. The term was introduced in the finalreport of the Commission for Environment andDevelopment (the Brundtland Commission). Thisphrase is empty, however, unless it is defined interms of operationally measurable properties,desired goals and relevant data. There are altern-atives to choosing ecosystems as the basis forenvironmental typology (Mackay & Paterson1982; O’Neill et al. 1982; Cairns & Pratt 1987).

1600–2000

1200–1600

800–1200

400–800

200–400

0–200

LAKE EKOLN

Mercury content(ng g−1 ds)

VardsatraBay

RiverFyris

GranebergBay

Hg content(ng g−1 ds)

0 4000

20

40

60

80

1970

1930

1890

1850

20

40

60

1970

1890

0 400



0

20

40

60

80

100

1970

1930

1890

1850

0 800 1600

N

0 1.5 km

Fig. 4.16 Sediments as a historical archiveillustrating both the areal and the temporalperspectives of ongoing changes, hereexemplified by the mercury contamination inLake Ekoln, Sweden. The sediment age has beendetermined from data on sedimentation insediment traps. The main tributary to this lake is River Fyris and mercury is transported to thelake from different sources (hospitals, dentists,laboratories, etc.) mainly from the city ofUppsala. (Modified from Håkanson & Jansson 1983.)

ESC04 9/5/06 3:03 PM Page 138


Instead, one might use different geographicalareas or different media such as air, water andsoil. There is, however, a clear international trendtowards consideration of the ‘health’ of differ-ent ecosystems (Bailey et al. 1985).

The majority of threats involve chemicals. Aset of ecological effect variables is expected toreflect such threats and the extent to which theyaffect the ecosystem. There is also a differencebetween biological effects for individual ani-mals or organs and ecological effects for entireecosystems. Practically useful, operational effectvariables should be measurable, preferably sim-ply and inexpensively, clearly interpretable andpredictable by validated quantitative models,internationally applicable, relevant for the givenenvironmental threat and representative for thegiven ecosystem.

Ecosystem indices would have the advantageof expressing the environmental status simply,but they simultaneously pose problems in that a great deal of valuable information is lost inaggregating the individual measures. This dis-advantage is reduced if it is known exactly what an index represents, and if these individual com-ponents can be accessed as required. Ideally, thesame basic framework would be used at both thenational and regional scales. As problems andpriorities cannot be completely congruent at different levels, however, the framework may beadapted to the different requirements of differentlevels. The national level may address large-scalethreats, perhaps originating outside the country,such as acidification of soil and water, whereasthe region can address more local problems, suchas the eutrophication of lakes (Case Study 4.3).

Case study 4.3 Impact of eutrophication/oligotrophication: Lake Miastro, Belarus

The following case study concerns Lake Miastro, Belarus (see Case Table 4.2 for lake data; for further details about the scenario, see Håkanson & Boulion 2002) and uses the well testedlake foodweb model, LakeWeb, for the simulations. The aim here is to illustrate both theeffect–load–sensitivity analysis and how changes in a critical load factor (the tributary con-centration of total phosphorus, TP) will cause, first, changes in lake phosphorus concentrations,in concentrations of suspended particulate matter (SPM) and in sedimentation. It also showshow such changes will influence target operational effect variables in lake management, such asSecchi depth (a measure of water clarity) and algal volume (a measure of lake trophic level). In1990 a drastic and sudden change in agricultural land-use practices occurred in the catchmentarea of this lake. The use of imported fertilizers was stopped as a result of political changesrelated to the fall of the former Soviet Union.

The presuppositions for this scenario are given in Case Fig. 4.3a, with the ‘modelled values’curve illustrating the sudden change in tributary total phosphorus (TP) concentrations in week261 (= January 1990; week 1 is the first week of 1985 and the simulation covers a period of 10 yr). There are good empirical data for this scenario (from Professor Alexander Ostapenia,Belarus State University, Minsk) giving mean characteristic monthly values for the period 1985to 1989 for lake TP concentrations (see Case Fig. 4.3b) and TP concentrations in sediments(Case Fig. 4.3e). One can first note the good correspondence between modelled values of lakeand sediment TP concentrations and empirical data. The following question is addressed in thisscenario (from Case Fig. 4.3b), how will the changes in TP concentrations in the lake influenceimportant sedimentological variables?

From Case Fig. 4.3c, one can note that the oligotrophication would likely imply a decrease insuspended particulate matter (SPM) concentrations because the algal volume would go down(Case Fig. 4.3g). This would also mean that sedimentation of matter decreases (Case Fig. 4.3d)

ESC04 9/5/06 3:03 PM Page 139

140 LARS HÅKANSON

1 131 261 391 521Weeks

1 131 261 391 521Weeks

1 131 261 391 521Weeks

0

50

100

Tot

al p

hosp

horu

sco

ncen

trat

ion

in la

ke (

µg L

−1)

0

10

20

0

0.5

1

Sed

imen

tatio

n (c

m y

r−1)

0

1.5

3

January 1990

Empirical data

Modelled values

SP

M (

mg

L−1)

Tot

al p

hosp

horu

s co

ncen

trat

ions

in A

-sed

imen

ts (

mg

g−1 d

w)

Modelled values Empirical maximum data

Empirical minimum data

(c)

(a)

(b)

(d)

0

50

100

Tot

al p

hosp

horu

s co

ncen

trat

ion

in tr

ibut

arie

s (µ

g L−1

)

Driving variable

0

1

2

Sec

chi d

epth

(m

)

Empirical data

Modelled values

1 131 261 391 521Weeks

1 131 261 391 521Weeks

1 131 261 391 521Weeks

1 131 261 391 521Weeks

Alg

al v

olum

e (m

m3

L−1)

10

5

0

Alarm limit

Critical limit

(e) (f)

(g)

Case Fig. 4.3 Case-study on eutrophication/oligotrophication in Lake Miastro, Belarus. There was a drastic reduction in theuse of fertilizers in argriculture in 1990. This has influenced the river concentration of phosphorus, as shown in (a), the drivingvariable in this scenario. It can then be asked how this will influence: lake concentrations of phosphorus (b), concentrations of suspended particulate matter (SPM; (c)), sedimentation of matter (d), phosphorus concentrations in A-sediment (e) andoperational variables of ecosystem effect, such as the Secchi depth (f ) and the algal volume (g).

ESC04 9/5/06 3:03 PM Page 140


An environmental index must be based on thestatus of some crucial characteristics of chosenecosystem types. These are the six basic eco-system types.1 Forests2 Agricultural land3 Natural land4 Freshwater5 Coastal areas6 Urban areasGenerally, it is extremely difficult to distinguishcause and effect in natural ecosystems. One can-not base an environmental consequence analysison a full understanding of the ecosystem. Incomplex ecosystems ‘understanding’ at one scale(e.g. the ecosystem scale) is generally related to processes and mechanisms at the next lowerscale (e.g. the scale of individual animals and/orplants), and the explanation of phenomena atthis scale is related to processes and mechanismsat the next lower scale (e.g. the scale of the organ),and so on down to the level of the atom andbeyond. In environmental management, a balancemust often be found between answering inter-esting, often important, questions of understand-ing, and delivering a practical tool to society. If an ecosystem index were based on a causalanalysis of what takes place at the cellular level,then at levels involving organs, individuals, populations, and finally at the ecosystem level, it

would be an eternity before the index could bedeveloped. For the foreseeable future, ecosystemindices are more likely to be based on practicalconsiderations of predictive power and samplingease, rather than full causal priority.

What, then, should the strategy be for develop-ing an ecosystem index? The first problem is thateach ecosystem type, for example fresh waters,is not a single entity. It consists of many sub-ecosystems (Fig. 4.17). A general resolution aboutthe basis of this approach is probably impossible,but questions about the appropriate hierarchicallevel of analysis are relevant to specific threats.Figure 4.17 lists the 12 general environmentalthreats. If one starts with the threat to freshwaters, it is clear that contamination from, forexample, metals and radionuclides threatensfresh waters and that these threats might bemanifested in, for example, reduced biologicaldiversity and contaminated fish.

4.5.2 Effect–load–sensitivity analysis

Elevated concentrations of contaminants thatcause no visible or measurable ecological effectswould generally be of less interest for practicalwater management, and for remedial strategies,in the situation faced today in ecosystems withmultiple threats. The aim of effect–load–sensitivityanalyses (ELS) and ELS models is to provide a

and the Secchi depth increases (Case Fig. 4.3f). It is interesting to note there are no clearchanges in TP concentrations in sediment (Case Fig. 4.3e) because phosphorus is a very mobileelement in lake sediments and the phosphorus concentration in sediments depends more on thesediment chemical ‘climate’, i.e. on redox conditions, than on deposition (see Table 4.5).

The drastic reduction in phosphorus loading to this lake can have both positive and negativeconsequences for the lake ecosystem depending on management objectives and criteria. From asedimentological perspective, one can conclude that the SPM concentrations probably willdecrease, the water clarity will be higher and the deposition of matter also will decrease. Theseare signs of oligotrophication.

Relevant reading

Håkanson, L. & Boulion, V. (2002) The Lake Foodweb – Modelling Predation and Abiotic/Biotic Interactions.Backhuys Publishers, Leiden, 344 pp.

ESC04 9/5/06 3:03 PM Page 141

142 LARS HÅKANSON

tool for quantitative predictions which relateoperationally defined ecological effects to com-patible load and sensitivity variables (Håkanson1999).

Differential equations and mass-balancemodels are often used to handle fluxes (e.g. g X yr−1), amounts (g X) and concentrations (g X m−3) of all types of materials (such as gases,carbon, toxins and nutrients), but not ecosystemeffect variables (E). Statistical methods, such asregressions based on empirical data, are generallynecessary to relate concentrations of chemicalsto effect variables (E). In theory, both thesemodel approaches (see Fig. 4.18) may be usedfor the ELS models, provided that at least oneoperationally defined ecological effect variablerelevant for the load variable(s) in question isincluded in the model. Ideally, the E variable

should express the reproduction, abundance,mass or status of defined functional organisms(preferably fish at the top trophic level) thatcharacterize the given ecosystem. Such ideal effectvariables cannot normally be predicted with dif-ferential equations and mass-balance models. Ifthe more ideal effect variables cannot be deter-mined, then in practice one has to do the secondbest and define operational effect variables, suchas toxic concentrations in fish eaten by humans,the oxygen concentration in the deep-waterzone and the Secchi depth (see Table 4.10). Thesensitivity variables express how different lakecharacteristics (such as pH, colour, total-P con-centrations and lake mean depth) regulate the‘road between load and effect’. One and the sameload will cause different ecosystem effects in lakesof different sensitivities. For example, in lakes

Clim

ate

chan

ge

Thi

nnin

g of

atm

osph

eric

ozo

ne la

yer

Incr

ease

d le

vels

of o

zone

at t

he s

oil s

urfa

ce

Met

al/r

adio

isot

ope

cont

amin

atio

n

Aci

dific

atio

n

Eut

roph

icat

ion

Con

tam

inat

ion

by to

xic

orga

nic

com

poun

ds

Hea

lth e

ffect

s

Alte

red

land

scap

es in

pro

tect

ed a

reas

Dre

crea

sed

biol

ogic

al d

iver

sity

Intr

oduc

tion

of n

ew a

nd a

lien

spec

ies

Exp

loita

tion

of n

atur

al r

esou

rces

1 2 3 4 5 6 7 8 9 10 11 12

ELS models availablefor 137Cs and Hg

Oligotrophic Mesotrophic EutrophicTarget ecosystemfor metalcontamination

In-flow areas Out-flow areas Lakes Rivers

Freshwater

Soil waterSurface water

Ecosystem

12 environmentalthreats

Fig. 4.17 The fresh water ecosystem may bedivided into several subecosystems (such asoligotrophic, mesotrophic and eutrophic lakes)where different key organisms prevail. As part of this differentiation process, one must decide whichsubsystem should be used as the target ecosystemfor a given chemical threat, which target organisms,functional groups or effect variables should be usedfor a given threat, and what load and sensitivityvariables should be used relative to a given effect variable. The target ecosystem for metals (such as Hg) and radionuclides (such as 137Cs) is low-productive (oligotrophic) lakes (shaded in this figure): ELS, effect–load–sensitivity. (From Håkanson & Peters 1995.)

ESC04 9/5/06 3:03 PM Page 142


with low pH and low bioproduction, a givenload of mercury will cause significantly higherbiouptake of mercury and hence also highermercury levels in fish, than in lakes with highpH and high bioproduction. There are manydocumented cases like this (see Håkanson 1999).One classic way to develop ELS models is to usedynamic mass-balance models to handle con-centrations of pollutants and empirical models(such as regressions) to link these concentrationsto the operational effect variables.

4.6 SUMMARY

Lake sediments are an important, integral partof the lake ecosystem. They reflect changes inland-use and lake characteristics and are oftenregarded as a historical archive. They also affectthe present structure and function of lake eco-systems. Suspended particulate matter (SPM) in

lakes mainly comes from tributaries (alloch-thonous matter), from living and dead matterproduced in the lake (autochthonous matter) orfrom resuspended materials. The SPM regulatesthe transport of all types of water pollutants indissolved and particulate phases; it regulateswater clarity and the depth of the photic zone,and hence also primary and secondary produc-tion; it regulates bacterioplankton productionand biomass, and hence also mineralization,oxygen consumption and oxygen concentrations;and it regulates sedimentation, and hence alsothe use of sediments as a historical archive, forexample of water pollutants. These matters arediscussed in this chapter. The aim of the infor-mation given is to structure existing knowledgeon the factors regulating variations among andwithin aquatic systems of suspended particulatematter in a rational manner. This knowledge isfundamental for an understanding of the func-tion and structure of aquatic systems.

Compartmentload

Inputload

Q . Cin Q . C

Ecological effectvariable

KT . C . V

Changes in Cin means

change in C

Mass-balance model Effect–load–sensitivity model

Amounts, fluxes andconcentrations

Ecological effects for entire ecosystems

Environmentalsensitivityvariableor function

Ecosystem load = dosevariable or function

Change in sensitivity may change the ecologicaleffect variable

Change in load, Cin or C,may change the ecologicaleffect variable

Fig. 4.18 Illustration of the fundamental difference between dynamic, mass-balance models and effect–load–sensitivity (ELS) models.The three wheels indicate that by means of remedial measures one may reduce the load variable in dynamic models and the load and thesensitivity variables in ELS models. KT is the sedimentation rate. (From Håkanson & Peters 1995.)

ESC04 9/5/06 3:03 PM Page 143

5.1 INTRODUCTION

Within arid landscapes sediment transfer beginswith the breakdown of bedrock to transportablematerial. This can be achieved by weatheringprocesses, which in arid areas are dominated bymoisture and salt or insolation. Owing to the roleof gravity in the transportation processes, mostsource areas will be located within the uplandareas of the arid landscape. These areas are typically dominated by zones where the erosivecapability (erosivity) of water is high, such as steepslopes. Other sediment source areas are domin-ated by zones with sediment that is susceptibleto erosion owing to its weak material charac-teristics (its ‘erodibility’), such as exposed lakesediments (Fig. 5.1). These available materialswill be transported by wind or water to (i) areasof temporary storage within transport zones, suchas rivers, or (ii) a more permanent storage area ofnet sediment accumulation, such as an untrenchedalluvial fan, a sand sea or a lake. These sedimentstores will themselves be susceptible to recycling

5Arid environments

Anne Mather

within the sediment transfer system (Fig. 5.1).This chapter examines the processes involved inthis transfer of sediment within the arid envir-onment, and the environmental hazards thatoccur en route. It provides an insight into therecent sedimentary deposits of arid environmentsand how they are affected by changing externalenvironmental controls.

5.1.1 Definition of arid environments

In this chapter arid environments will be con-sidered to embrace regions described both as‘drylands’ and ‘deserts’. Deserts are more qual-itatively defined by a range of physical criteria,whereas drylands are quantitatively defined and classified using meteorological data. As the definition of drylands is based on modernclimate data records, however, it is difficult to rigorously apply this definition to older historic to geological time-scales. As a resultboth terms (drylands and deserts) have tendedto be used interchangeably in the literature,

Lakes/playas/sabkhas

Ephemeral channels

Alluvial fans

Aeolian dunes

Slopes

Storedsediment

Water

Sediment storageSediment sources Sediment transport

Weathering

WindFig. 5.1 Simplifiedflow diagram ofsediment productionand routing in aridenvironments.

ESC05 9/9/06 2:42 PM Page 144

ARID ENVIRONMENTS 145

with deserts describing the hyper-arid to ariddefinitions of drylands. Thus, for the purpose of this chapter arid environments will be con-sidered to represent the more extreme drylands(hyper-arid to arid; 4% and 15% of the Earth’sland surface respectively), which can be recog-nized in the older landscape records, and thosedrier areas at greatest risk of degradation byhuman activity (semi-arid, representing 15% of the Earth’s land surface; Fig. 5.2a). A fullerdefinition of the terms ‘desert’ and ‘dryland’ is given below.

Deserts are inhospitable, barren, poorly veget-ated areas devoid of water. There are numerousscientific definitions of deserts available basedon a range of physical criteria from erosion to vegetation type (see Heathcote (1983) andThomas (1997a) for a summary). Deserts there-fore cover a range of climate categories, rangingfrom cold arid (Gobi Desert) to hot arid (centralSahara), and some deserts are so large they maycover several climate types – the Sahara includeshot (central Sahara), mild (southern Sahara), andcool (northern Sahara) climates. Most people

Hyper-arid

Arid

Semi-arid

0–20%

Over 30%

Rainfall variability

46

2

7

8 19

5

3

1234

56789

ArabianAtacamaAustralianChihuahua

Deserts GobiMojaveNamibSaharaThar

5.1

5.25.4

5.25.3

5.15.3

5.4

South East SpainSouth West USACentral & SE SpainCentral USA

Case Studies

(a)

(b)

Fig. 5.2 (a)The global distribution of arid environments and the location of the major deserts referred to in the text. (Modified fromGoudie & Wilkinson 1978.) (b) Rainfall variability and locations of Case Studies used in this chapter. (Modified from Dregne 1983.)

ESC05 9/9/06 2:42 PM Page 145

146 ANNE MATHER

associate deserts with areas dominated by exten-sive sand dune systems (sand seas). In reality,however, these landscape elements are relativelyminor compared with other elements such asmountainous areas (Fig. 5.3).

Drylands are classified quantitatively intohyper-arid, arid, semi-arid and dry-subhumidusing meteorological data. These classificationsare defined using the balance of moisture inputs(precipitation) and losses (evapotranspiration)expressed as an aridity or moisture index. Thisindex is typically expressed as the ratio of P(precipitation) to PET (potential evapotranspi-ration). Classifications vary, however, depend-ing on the data used (Table 5.1). For exampleMeigs (1953) used aggregated monthly toannual moisture surplus and deficit data to rep-resent P and the Thornthwaite approach to calculate PET. These data are not widely avail-able from arid regions. More recent definitionstherefore have tended to use annual precipita-tion for P and the Penman method of calculatingPET, which utilizes an understanding of the

diffusion of water vapour and aerodynamicfunctions, data which are universally avail-able. To further facilitate a comparable globalcoverage the data sets used in the most recentdefinitions also tend to be time-bounded toavoid data bias from regional data sets wherethe records are less continuous. The time periodspans some 30 years to take account of inter-annual and interdecadal variations in climate.Variation in the data sets used means thatfigures on the modern extent of drylands rangefrom 26.3 to 47.2% of global land area,depending on the classification used (Thomas1997a).

5.1.2 Causes of aridity

5.1.2.1 Climate

Arid environments are typified by variable rain-fall (characteristically more than 30% of themean, Fig. 5.2b). This rainfall will range from a number of low-intensity events, which may

0%

20%

40%

60%

80%

100%

SW USA Sahara Arabia Libyan Desert

Mountain

Desert flat

Regions bordering through flowing rivers

Fan and bajada

Badland

Playa

Bedrock field

Dry wash

Aeolian dune

Volcanic features

Fig. 5.3 The percentage oflandscape units found within fourtypical deserts. (Data from Clementset al. 1957.)

ESC05 9/9/06 2:42 PM Page 146


occur over several days but generate only limitedrunoff, to high-intensity rainstorm events, whichgenerate flash flood events but may occur onlytwo or three times in a 100-year period. Regionsof such restricted rainfall are most commonlyrelated to four main individual climate causes,which may be related in part to tectonics (seesection 5.1.2.2). In addition the individual causeslisted below may interact with each other, rein-forcing conditions of aridity.1 Global atmospheric circulation. Areas dom-inated by high-pressure cells and hot (such asthe tropics of Cancer and Capricorn) or cold drysubsiding air (the North and South Poles) willhave limited rainfall capability.2 Continentality. In large continents (e.g. theinterior of Asia) available moisture may havebeen precipitated out in the more coastal areasof the landmasses.3 Rain-shadow created by high mountain areas.In coastal ranges such as the Rockies, USA, moistair derived from evaporation over the ocean

rises and is cooled adiabatically. Thus vapoursaturation is reached quickly and precipitated asrain or snow. The desiccated air will then descendacross the mountain range, becoming warmerand drier into the area of rain shadow.4 Cold upwelling ocean currents. The mostprominent examples are the Humbolt current ofthe Atacama Desert and the Benguela current of the Namib Desert. Both these ocean currentsare derived from the colder, southern polar latitudes and flow northwards. Their associatedcool moist ocean air reaches relatively warmerland, where its relative humidity decreases andits capacity to absorb moisture is increased. Thusthe oceanic air mass tends to desiccate thesecoastlines. The main source of moisture in suchareas is coastal fogs such as the ‘camanchaca’ ofthe Atacama.

Moisture in arid areas will not be solelysourced from rainfall events. Coastal fog may be important in arid regions that border oceans(e.g. the Atacama Desert of South America and

Table 5.1 Quantitative classifications of arid zones. (Based on Thomas 1997a.)

Classification and use

UNESCO 1979

Grove 1997

UN 1977

UNEP 1992

Data used

Uses Meigs’ (1953) classificationtogether with Thornthwaite’s (1948)indices of moisture availability (Im)Im = (100S − 60D)/PETwhere S is the moisture surplus, M themoisture deficit and PET is potentialevapotranspiration. S and M areaggregated on an annual basis frommonthly data, taking stored soilmoisture into accountMeigs 1953 classification, with rainfallfigures

P/PET index: where P is annualprecipitation and PET is potentialevapotranspirationAridity index (AI)AI = P/PETwhere P is annual precipitation andPET is potential evapotranspiration.PET is calculated using theThornthwaite method. Data are takenfrom time-bounded study

Classification

Semi-aridAridHyper-arid

Semi-arid = 200–500 mm yr−1

Arid = 25–200 mm yr−1

Hyper-arid = < 25 mm yr−1

Semi-arid 0.20 ≥ AI < 0.50Arid 0.03 ≥ AI < 0.20Hyper-arid AI < 0.03Dry-subhumid 0.50 ≥ AI < 0.65Semi-arid 0.20 ≥ AI < 0.50Arid 0.05 ≥ AI < 0.20Hyper-arid AI < 0.05

Comments

Does not include areas too cold for crops (e.g. polar deserts)

Approximate rainfallfigures. Figures availableover widely varied time-scales spatiallyPET calculated byPenmans formula usingdata not globally availableIncludes dry subhumidareas

ESC05 9/9/06 2:42 PM Page 147

148 ANNE MATHER

the Namib Desert of southern Africa; Fig. 5.2).Gobabeb research station (Namibia) data recordsshow that coastal fog is a significant contributorof moisture, even though the research station is some 60 km inland and 408 m a.s.l. Here fogmoisture may well exceed rainfall in some years.Another significant source of moisture may bedew. Fewer records on dew are available butdata from the Negev, Israel, suggest that dewoccurs on an average of 195 nights per year atthe ground surface, and provides 33 mm yr−1 ofmoisture, although only 0.35 mm for any onedew night (Evanari et al. 1982).

5.1.2.2 Tectonics

Arid environments can be found in a range oftectonic settings. These include currently stableintraplate settings such as cratons (e.g. theAustralian Desert) and passive margins (e.g. the Namib). Arid regions can also be found inextensional tectonic settings such as the Mojaveand Chihuahua deserts, and compressional con-tinental margins such as the Atacama and Thardeserts. Some deserts, such as the Sahara, are solarge that they may include several tectonic set-tings (the Sahara spans all of the above settings).

Despite the apparent lack of correlation betweentectonic settings and arid region occurrence, tectonics are a major control on the location andpersistence of arid environments. Tectonics con-trol the regional and global topography of thecontinents, which in turn has significant effects onthe local climate. The main impact of tectonicsas a control on the development of arid areas isthrough the two processes listed below.1 Orogeny. Tectonics can generate mountains,which create local rain shadows. On a largerscale, large orogenic belts such as the Himalayascan divert major air circulation, leading to aridity. The associated uplift will also lead to anincrease in the continentality of an area.2 Continental drift. Tectonics can affect latitudeover many millennia through continental drift(consider for example Britain which was muchcloser to the Equator 250 Ma in the Permo-Triassic, and subject to desert conditions). Platetectonics can also lead to the closure of oceanicgateways, which can affect oceanic currents and

thus coastal climates. It has been postulated thatthe closure of one such gateway reinforced thehyper-aridity of the Atacama Desert of SouthAmerica (Hartley 2003).

5.1.2.3 Anthropogenic agents

Humans have been credited with forming andexpanding many modern deserts (Ehrlich &Erlich 1970). The expansion of desert areasthrough human intervention is known as‘desertification’ (sensu Mabutt 1978, 1985).This process, and its definition, however, hasbecome increasingly disputed (e.g. Williams1994). There is no doubt that mismanagementof land through overgrazing and soil saliniza-tion has occurred since the Holocene increase in human population. The human population is estimated to have increased from 107 at thestart of the Holocene Epoch (10 ka) to 108 by 5 ka and 109 by 2 ka (May 1978). Typically,however, the increase in desert areas is tem-porary and associated with intervals of drought,which in most cases return to their natural non-desert habit once the drought has ended andrecovery set in. In 1992 the Rio Earth Summitthus defined desertification as ‘land degrada-tion in arid, semi-arid and dry subhumid areas[‘susceptible drylands’] resulting from variousfactors, including climatic variations and humanactions’ (UNEP 1994).

5.1.3 Variability of arid environments through time

The geological record of aeolian dunes and eva-porite sediments indicates that arid areas havebeen present since the Precambrian (Glennie1987). The Atacama is probably the oldestextant desert in the world (Hartley et al. 2005),dating back to the Jurassic (some 150 Ma), butmost deserts have retained their position over thelast 2 Ma. Although the locations have remainedrelatively fixed over these geological time-scales,the actual surface area has changed. For example,some of the drier areas of Africa appear to havebeen expanding as a result of persistent droughts.Typically, however, these fluctuations are tem-porary. Quaternary records from tropical and

ESC05 9/9/06 2:42 PM Page 148


subtropical regions, however, do indicate that thespatial distribution of arid regions has changedover this time frame (Fig. 5.4).

5.2 SEDIMENT SOURCES AND TRANSPORT

This section will examine how the natural aridlandscape functions from source to storage, following the dominant transport processes and in situ modification (weathering) that can occurafter deposition. Weathering is highly variable,depending on localized differences in climate(mainly temperature and humidity), geology(mainly lithology and structure) and inheritedweathering artefacts (Smith & Warke 1997).Weathering is important as it weakens the bed-rock ready for erosion and transport.

5.2.1 Significance of weathering in arid regions

Arid environments are typified by weathering-limited sediment supply. Exposed bedrock iscommon and the lithological and petrographiccharacteristics of the geology are more importantthan in humid environments in controlling local-

ized rates of weathering and sediment production.This also means that any inherited weaknesses,for example developed from weathering in pastclimate regimes, become more significant asthese features are key to the in situ weatheringdevelopment and thus, ultimately, erodibility ofthe sediments.

5.2.1.1 Insolation

Large temperature ranges (which can exceed50°C, Goudie 1997) associated with the diurnaltemperatures of deserts have been attributedwith expansion and contraction of rocks in situ.Early workers suggested that this was the mainfactor behind rock breakdown. Experimentalwork, however, has failed to reproduce the samedegree of weathering under laboratory condi-tions, leading to much debate about the possiblerole of insolation (see Goudie (1997) and dis-cussions therein).

5.2.1.2 Moisture

Although by their definition arid areas have limited water availability from rainfall, moisture

50 45 40 35 30 25Years BP × 103

20 15 10 5 0

Africa– northof theEquator

Maghreb (Morocco,N. Algeria, Tunisia)

Northern Sahara

Eastern Sahara

Sudan

Chad

Middle Kalahari

Southern Kalahari

Southern/interiorAustralia

Hyper-arid – aridLow lakes, activedunes, etc.

Arid – Semi-aridEphemeral lakesand rivers. Localdune activity

Humid – Sub-humidHigh lakes andrivers. Cave sinterdevelopment, etc.

North-east Australia

Thar Desert

SouthernAfrica

Australia

India

Fig. 5.4 A simplified chronology of late Quaternary expansion and contraction of tropical and subtropical arid zones derived fromgeomorphological and sedimentological data. Blank areas in the bars represent periods of uncertain dominant climate (due to climaticinstability or a lack of data). (From Thomas 1989.)

ESC05 9/9/06 2:42 PM Page 149

150 ANNE MATHER

from other sources may be significant (section5.1.2.1). Such sources of moisture tend to bealkali as they are sourced from salt-rich waterssuch as the sea or salt lakes. In the latter casespH may be as high as 9.9 (Van Gulu, Turkey;Goudie 1997). This high pH greatly increasessilica mobility during rock weathering. Wheremoisture is trapped within rocks the thermalexpansivity of the rock (Goudie 1997) is in-creased. Heating of 10–50°C can develop 250atmospheres of tensile strength (Winkler 1977),thus exacerbating breakdown of the rock.

5.2.1.3 Salt weathering

In arid environments salt can be sourced from anumber of potential areas. These may include:1 sea water or relict sea water which may befound in bodies of water (e.g. lakes) formerlyconnected to the sea;2 ‘cyclic salts’ derived from atmospheric inputssuch as dust and rainfall or volcanic emissions;3 the release of salts through rock weathering,especially where the bedrock is evaporitic in origin (e.g. halite).These abundant salts in arid environments meanthat salt weathering is not only key in weathering,but also creates problems in the built environ-ment (section 5.6.3).

Weakening of sediment by salt weatheringpredisposes many types of sediment in arid areasto deflation. In some cases the hollows gener-ated by deflation attract grazing animals, whichin turn increase erosion (Goudie 1989). Erosionof the pans may be limited at depth by ground-water (Goudie & Wells 1995), which will alsocontribute to the salt weathering. Salt weather-ing can occur through a number of processes.1 Salt crystal growth. This may result fromchanges in solubility with temperature, evapora-tional concentration of solutions, and mixing of salt solutions with the same ion (see Goudie1997).2 Hydration. Some salts will hydrate and rehydrate in response to changes in temperatureand humidity. As salt changes to its hydratedform it takes up water, increasing its volume.Some of the largest hydration pressures occur

during anhydrite (dehydrated calcium sulphate)to gypsum (hydrated calcium sulphate) trans-formation (Winkler & Wilhelm 1970).3 Thermal expansion. When halite (sodiumchloride) is heated from 0°C to 60°C it expandsby 0.5%, whereas granite minerals only expandby up to 0.2%. This differential expansion cancontribute to rock disintegration (Goudie 1997).

5.2.2 Zones of net erosion

Within arid environments sediment productionwill be dominated by source areas susceptible toerosion. These most commonly include areas thatgenerate conditions of high erosivity (slopes) orareas of high erodibility reflecting lithologies lessresistant to erosion, such as lake-bed sediments.

5.2.2.1 Slopes

Many arid areas are associated with landformssuch as pediments, cuestas and mesas that areprotected by a caprock. This caprock may bepart of the geological sequence, or it could be a crust developed in situ (section 5.3.5). What-ever the cause of the crust, caprock failure is not uncommon, producing slope talus and apotential sediment supply. Rates of scarp retreat,and thus sediment production, vary accord-ing to the local balance between the geologicalcharacteristics and the climate. Lower rates tend to be reported from hyper-arid regions, for example 100 mm kyr−1 from limestone hasbeen reported from the Sinai, Israel (Yair &Gerson 1974). However, the overall control onrates of retreat is local lithology. This variesgreatly between different lithologies, with thehighest retreat rates reported from conglo-merates (6700 mm kyr−1, Lucchitta 1975) andsandstone (500–6700 mm kyr−1, Schmidt 1980,1989). Similarly retreat rates within the sameoverall lithology may vary dramatically as a func-tion of microlithological variations in the caprock,and overall stratigraphy. For example, limestonescarp-retreat rates of 100–2000 mm kyr−1 havebeen reported in hyper-arid areas (Sinai, Israel;Yair & Gerson 1974) and 160–400 mm insemi-arid areas (Arizona, USA; Cole & Mayer

ESC05 9/9/06 2:42 PM Page 150


1982; Young 1985). Talus will accumulate as a function of mechanical failure of the caprockand fall impact. Typically coarse talus tends to be associated with lithologies that break intolarge chunks, i.e. lithologies composed of largecomponent parts, such as conglomerates, orlithologies of greater mechanical strength, suchas limestones. It is not uncommon for poorlycemented sandstones, such as aeolianites, to beassociated with no talus at all, the sediment beingcarried away by contemporaneous aeolian pro-cesses (Schumm & Chorley 1966).

Once the talus slopes have been created theymay generate their own runoff (Yair & Lavee1976) in rainstorm events. This typically occursclose (within 10 cm) to the surface. Depend-ing on the balance between talus supply andremoval, the talus may be eroded to expose thearea previously protected by them, revealing talus‘flat irons’. This is common in some parts ofUtah, USA. Alternatively in hyper-arid settingsthe talus may form a ‘fossil’ accumulation in themodern landscape.

5.2.2.2 Exposed lake sediments

Lake-bed deposits in arid regions range in thickness from a thin veneer to hundreds ofmetres. As water bodies expand and contract in response to changes in the water balance, soformer lake-bed sediments may become exposedto further erosion by surface processes such asmass wasting or gully erosion, being cannibalizedinto the lower lake levels.

Lake sediments contain a range of grain sizes,depending on the extent and depth of the lakesystem, and the characteristics of the surround-ing catchments feeding the lake with water andsediment. Thus lake sediments may have beensourced originally from terrigenous, allochthon-ous sediments external to the lake water body.These clastic sediments are typically transportedby runoff and river flow and thus can contain arange of grain sizes from clay to cobble. Othersources of lake sediment are autochthonous andoriginate within or beneath the water column orwithin the lake-bed sediments after deposition.These sediments are typically fine-grained, but

often well-cemented carbonates. The finest ofthese lake sediments will be prone to dust ero-sion (section 5.2.4). Where this occurs sedimentwill be exported out of the basin system. Onesuch relatively recent example of this is OwensLake, California, which is recorded in somedetail in Reisner (1986) and Knudson (1991).Owens Lake was a relatively small (110 km2)water body that was located to the south ofOwens Valley (Case Fig. 5.2A). Early Europeansettlers were attracted to the area and it becameimportant for agriculture, with irrigation farm-ing introduced. In the 1900s, however, pressureon available water resources in the growingconurbation of Los Angeles was high. This ledto the purchasing of the water rights of this area,by members of the Los Angeles Water Companyposing as cattle ranchers. In 1913 many of thelocal streams that fed the lake were diverted intothe Los Angeles Aqueduct. In 1927 the lake wasnothing more than a small pond. The exposeddry, fine-grained playa sediments contain halite,trona, thenardite and mirabilite and are suscept-ible to erosion by strong summer and winterwinds. Dust derived from this 110 km2 areaaccounts for 1% of the total dust production in the USA every year. In February 1989 a con-centration of 1861 μg m−3 was recorded, whichwas 37 times higher than the health standard inthe State of California.

Where lake sediments become elevated inrespect to the main lake level as a function oftectonics, then mechanical erosion of all gradesof lake sediment material becomes possible fromgully erosion. This is currently occurring in LakeBurdur in central Turkey, where tectonicallyelevated Pliocene lake sediments are being erodedby gullying, and redeposited in the smaller, andconstantly shrinking, modern lake system.

5.2.3 Sediment transport by water

Channels within arid areas may be one of two main types: (i) perennial or (ii) ephemeral.Perennial channels have flow for most of theyear and exist within arid areas where riverswithin the arid basins are sourced by externalinputs from mountainous regions. For example,

ESC05 9/9/06 2:42 PM Page 151

152 ANNE MATHER

in the Euphrates 90% of the total runoff comesfrom the mountains of north-eastern Turkey,with virtually zero runoff from areas such asIraq (Beaumont 1989). Similarly the Nile flowsfrom highlands in Ethiopia, which supply mostof its water, but most of its course is through the arid lands of Egypt. Before the constructionof the Aswan dam the Nile annual dischargeranged from a low of 1000 cumecs (cubic metresof water per second) in spring to 11,000 cumecsin late summer/early autumn (Beaumont 1989).Perennial rivers in general form the focus ofChapter 3, so will not be discussed further in thissection. Ephemeral channels dominate in aridzones and thus form the focus of the rest of thissection. These channels are only occupied bywater following a rainstorm as they are mainlysupplied by storm runoff. They tend to show lessregular discharge patterns than perennial rivers,and suffer large transmission losses downstream.

5.2.3.1 Flow characteristics of ephemeral channels

Arid river systems cover a diversity of forms on a variety of scales. Cooper Creek, Australia,for example, is the largest internally drainingcatchment in the world (1.3 × 106 km2; Knighton& Nanson 1997) and lies entirely within an arid zone. Owing to the hydrological character-istics of the slopes feeding ephemeral channels(section 5.4.3), runoff is a typically more signific-ant component in streamflow than in more humidareas, where throughflow and groundwater maydominate. For example, in New South Wales,Australia, only 16 mm of rainfall is required toproduce runoff in the arid west, whereas 35 mmis required in the more humid east (Cordery et al.1983). In Australia large, tropical monsoonalstorms can penetrate large distances inland of thearid zones. In most arid systems, however, rain-fall is derived from small, often convectionalstorm cells of less than 10–14 km width (Renard& Keppel 1966; Diskin & Lane 1972). Eventhough these may activate only a small propor-tion of a larger drainage basin, large floods canbe generated rapidly. In the USA the top 12largest floods recorded have all occurred withinarid/semi-arid areas (Costa 1987). So, although

mean annual runoff is lower for most ephemeralstreams, compared with perennial streams, peak-flood discharges are similar (Fig. 5.5).

Owing to the transport-limited nature of thesystem, arid-zone rivers tend to have high peakvalues of suspended sediment concentrationsand they are much more efficient transporters ofbedload material than humid systems. In partthis is reflected in the wide channels that typifyephemeral river systems (Fig. 5.6). Thus a uniqueaspect of ephemeral rivers is their multiple terrace and channel cross-section. Large eventscut the widest channel and upper ‘terraces’ of

10−2 10−1 100

102101100 103 104 105

Drainage area (km2)

100

10−1

101

102

Max

imum

flo

od d

isch

arge

(m

3 s−1

)M

ean

annu

al r

unof

f (1

06 m3 ) 103

100

101

102

103

104

101 102

PerennialEphemeral

(a)

(b)

Fig. 5.5 A comparison of perennial (Maryland) and ephemeral(California) channels. (a) Differences in mean annual runoff,with largest values associated with perennial rivers. (b) Ademonstration of the similarity in the size of peak flooddischarges generated from perennial and ephemeral drainageareas. (Adapted from Wolman & Gerson 1978.)

ESC05 9/9/06 2:42 PM Page 152


the outer channel. Successively more frequent,lower magnitude floods occupy the topogra-phically lower and narrower parts of the riverchannel (Fig. 5.7). Burkham (1972) recordedchanges in river width over a series of floods in the Gila River, Arizona (Fig. 5.8). The largestof these floods had a recurrence interval of 300years and increased channel width by 600%.Recovery here took some 50 years, but in morearid areas it can take much longer for vegetationto re-establish itself. Thus relationships basedon bankfull discharge (considered the dominantchannel forming discharge in perennial rivers bymany authors) are of less relevance to ephemeralarid river systems.

One of the main differences between peren-nial and ephemeral river systems is in the pro-pagation of the flood wave. For the reasons discussed above, the flood wave will be highlyvariable as each rainstorm event potentiallyactivates a different area of the catchment. Also the small size of the rain cells (< 14 km),typically distributed 40–60 km apart (Sharon1974), means that for increasingly larger basinsa smaller percentage of the catchment will remainactive, unless the region receives rainfall fromlarger weather systems (e.g. Australia alreadydiscussed). In addition to this the rain cells willmigrate as they precipitate. Thus, tributarieswithin catchments may be activated at differenttimes. Runoff generation typically takes only a few minutes in arid environments, with gully

systems activated in around 20 minutes but maintributaries may take hours, depending on thescale of the system. Schick (1988) suggests a pig-gyback contribution of individual tributariesgenerates the multipeak appearance of many aridflash floods. This spatial difference in rainfallmay be increased where topographic variance is maximized; for example, in basin and rangestyle topographies.

The flash-flood hydrograph is dominated by asteep rising limb. Normally the first observationis a flood bore. These vary in height. Hassan(1990) reported flood bores of 4–40 cm inheight, and moving at 0.2–1 m s−1, with velo-city increasing with bore height. It is notable ingravel-bed rivers, however, that preceding thesurface bore is a subsurface percolation of waterthrough the river-bed coarse material. Floodbores can arrive suddenly and wreak devastation(Hjalmarson 1984) as they are rapidly followedby the flood peak within 10–23 minutes (Renard& Keppel 1966; Schick 1970; Reid & Frostick1987). The flood peak often consists of severalspikes reflecting the temporal and spatial activa-tion of tributaries within the system (Fig. 5.9).Within a few minutes of the bore average streamvelocities may have risen to 3 m s−1, and water-stage height has been reported to rise by 0.25 mmin−1 (Reid et al. 1994). The resulting floodmay last only a few hours. This is a function ofthe lack of sustaining rainfall and thus surfacerunoff, combined with a lack of subsurface flow

10−2 10−1 1 10

HumidSemi-aridAridHyper-arid

102

Drainage area (km2)

103

103

102

10

Ave

rage

ban

kful

l w

idth

(m

)

1104 105 106 107

Fig. 5.6 The relationship between bankfull width and drainage area under different climatic environments. (After Wolman & Gerson 1978.)

ESC05 9/9/06 2:42 PM Page 153

154 ANNE MATHER

and high transmission losses through the bed ofthe river. Transmission losses will vary accord-ing to the characteristics of the channel-fill sediments, the length of the flood and the widthof the flood. These will vary downstream. Over-all water discharge will decrease downstream as a result of transmission losses. It is thus clearthat ephemeral streamflow is highly complexand variable in nature.

5.2.4 Sediment transport by wind

Although winds in arid environments are nostronger than those in other regions, the sparsevegetation protection combined with large periods of transport inactivity by water meanthat sediment supply is typically abundant andthus the winds carry more sediment than anyother geomorphological agent (Cooke et al.

(a)

(b)

Fig. 5.7 An ephemeral river bed (Rio Aguas, south-east Spain)displaying multiple channel sizes and levels associated with varyingmagnitudes of floods. Photographs(a) and (b) represent the same cross-section, occupied by a lowmagnitude, annual event in (b). The whole width of the channelvisible here was last activated by a 1 in 500 year storm event in 1973,documented in Thornes (1974).

ESC05 9/9/06 2:42 PM Page 154


for example, 10,000 km from the Gobi Desertto Alaska (Rahn et al. 1981). The major sourceof this dust is from the semi-arid areas of theworld, particularly those desert environmentscontaining floodplains, alluvial fans, wadis, salt-pans and former lake beds (Middleton 1997).

5.2.4.1 Transport zones

The key zones where sediment transport by winddominates are in sand seas, sheets and stringers,with highest rates in sheets and stringers. Theseact as part of local and regional-scale aeoliantransport zones from source areas to sinks(Lancaster 1995) and are found most commonlyin areas with mean rainfalls of < 150 mm yr−1

(Wilson 1973). In such environments vegetationeffects are limited. Transport rates in dunelesssand sheets and stringers have been quoted as highas 62.5–162.5 m3 per metre width per year inMauritania (Sarnthein & Walger 1974) whereasmobile barchan dunes are reported at 3.49 m3

per metre width per year (Thomas 1997b). Thesezones of transport reflect local topography, andmay be sinuous and river-like in form (Zimbelmanet al. 1995). Typically net accumulation of sandsand bedforms (sand dunes and seas) will occuronly where accumulation exceeds net transport.In cases where net transport exceeds accumula-tion then sand sheets and stringers will dominate.

1993). Aeolian dust (suspended particles typ-ically less than 50 μm) is significant in globaltransport of sediment and it has been demon-strated that on a global scale the quantities ofdust in motion are of the same order of magni-tude as the quantities of sediment carried byrivers (Livingstone & Warren 1996). In areas ofnet sediment deposition, rates of accumulationof aeolian-derived sediment have been recordedat up to 70 mm kyr−1 (Dunkerley & Brown 1997)and dust can be transported over large distances,

18800

100

200

300

Ave

rage

cha

nnel

wid

th (

m)

400

500

600

1900 1920Year

1940 1960

Recovery

8 major floods

Reach A

Reach B

Fig. 5.8 Historical changes in channel width (1875–1970) forthe Gila River, Arizona. (Adapted from Burkham 1972.)

Tributary

High-magnitude

event

Low-magnitude

event

Transmissionlosses

Tributary

Tributary

RechargeRecharge

Tributary

Dis

char

ge

Downstream

Fig. 5.9 Conceptual model of transmissionlosses and their relation to flow in an ephemeralchannel for storms of two different magnitudes.The shaded area represents discharge. Note theintermittent channel flow for the low-magnitudeevent, and continuous but variable discharge forthe high-magnitude event as a function oftributary input and transmission losses. (After Thornes 1977.)

ESC05 9/9/06 2:42 PM Page 155

156 ANNE MATHER

5.2.4.2 Transport processes

As with water, sediment movement will occuronly where the erosivity of the wind is greaterthan the erodibility of the surface materials. Theerodibility will be determined by a number offactors. Vegetation will enhance surface rough-ness and modify local wind-velocity profiles. Indune systems this can reduce ground-surfacevelocities by as much as 200% (Wiggs et al.1994) thus limiting entrainment and encour-aging deposition. Surface slope will have animpact on both the threshold of sediment move-ment and the rate of sediment transport, andmay be more important than initial theories sug-gest (Hardisty & Whitehouse 1988). Moistureincreases surface tensions in sediments, reduc-ing erodibility, although the impact is stronglyaffected by the grain-size characteristics of the sediment (McKenna-Neumann & Nickling1989). In the case of moisture the most rapidchanges for erodibility occur at about 8% mois-ture content. Thus near-surface groundwatertables may limit sand transport (Stokes 1968) as can periodic flooding, both of which are particularly common factors in playa, sabkhaand coastal sand-sea situations (Fryberger et al.1988). Moisture contents up to 14%, however,will have no real impact on sediment already in transport (Sarre 1989, 1990). Salt and algalcrusts (section 5.3.5) may also protect surfacesfrom erosion.

Once entrained the sediments may be trans-ported by a number of different modes. Suspen-sion applies to smaller particles (< 0.6 mm). Veryfine sediment may remain suspended for manydays and travel great distances. Creep involveslarger particles (0.5–2 mm), which roll as aresult of wind drag and the impact of saltatinggrains. It has been estimated that creep accountsfor 25% of aeolian bedload transport rate (Willets& Rice 1985). Reptation represents the transi-tional state between creep and saltation and isgenerated by the physical impact of high-velocitysaltating grains on near-surface grains. Saltationis the most researched mode of transport. Saltat-ing sediments (typically 0.06–0.5 mm) move ina number of steps using a ballistic trajectory.

Although individual saltating grains can reach 3 m into the air (Pye & Tsoar 1990), most (80%)of all such transport takes place within 2 cm of the ground surface (Butterfield 1991). Theimpact of saltating grains on the surface maylead to reptation, or, where sufficient momentumis transferred, may induce mass transport bysaltation in a cascading system (Nickling 1988).Transport will be dependent on a number of factors including grain shape. For example,Willets (1983) showed that platy grains have atendency for lower and longer trajectories thanmore spherical grains.

5.2.4.3 Landforms created by aeolian transport processes

Ventifacts are common in areas dominated bywind erosion, such as adjacent to slopes. Theyare generated mainly by saltating, rather thansuspended grain movement. The latter tend tobe swept around an object rather than impact-ing with it. Sand-grade material in saltation thusaccounts for most wind erosion on the wind-ward side. Smaller particles of dust, althoughhaving less impact energy, can, in sufficient load(Whitney & Dietrich 1973), erode on the leeside of objects, utilizing wind eddies (McCauleyet al. 1977, 1979). There is, however, much controversy on the dominance of dust versussuspended load and its role in abrasion (see discussion in Breed et al. 1997). Larger scaleversions of abrasion structures are known as‘yardangs’. These have the form of an invertedboat, with a high windward side and stream-lined lee side. They range in size from a fewmetres to several kilometres. The largest yardangson Earth are up to 30 km long and developed in the Tibesti Plateau of the central Sahara (Peel 1970).

5.3 SEDIMENT ACCUMULATION PROCESSES

Sediment deposition can occur temporarily withintransport zones, or more permanently in areasof net accumulation (sediment storage; Fig. 5.1).This section will examine the sediment accumu-lation processes that affect these zones.

ESC05 9/9/06 2:42 PM Page 156


5.3.1 Desert lakes, playas and sabkhas

Lake basins act as sediment sinks, particularly inarid environments. Many arid zone areas eitherlack integrated drainage systems or are dominatedby endoreic (internal) systems. Thus, topographiclows act as foci for runoff collection and associ-ated sediment deposition. For example, erosionhollows generated by deflation may occur wherecrusts (section 5.3.5) are broken in arid regions.These hollows can later act as collection areasfor runoff, forming playas. In most cases thesewater bodies tend to be supersaturated with salts(> 5000 mg L−1) and are ephemeral. The associ-ated landform is known as a ‘playa’ or ‘pan’. Theplayas and pans vary in size from a few squaremetres to several thousand square kilometres.Their spatial significance varies regionally, butaccounts for only 1% of drylands (Shaw &Thomas 1997). There are a wide range of ter-minologies used for the same feature, with theterm ‘pan’ used for small areas originating from geomorphological (e.g. a depression betweenalluvial fans) rather than geological processes(Goudie & Wells 1995) and ‘playa’ for a depres-sion with a saline surface (Rosen 1994).

Sediments trapped in pans and playas are typically fine-grained and represent the suspendedsediments of distal flood inputs from adjacent

catchments and fan environments. Fine mater-ials are also supplied by aeolian transport. Inaddition evaporitic deposits may accumulate,sourced by evaporation of flood water or ground-water sources and the resulting concentrationand accumulation of salts (Fig. 5.10). The flatmorphology typically associated with playas is a function of infrequent inundation by water,which evens out the microtopography through a combination of deposition and dissolution.Playas with a highly irregular topography (e.g.from salt growth or sand-dune development inthe centre) suggest extremely infrequent inunda-tion by water, although an uneven topographyfrom alluvial fans and drainage courses aroundthe margins is not unusual (Fig. 5.10). To main-tain a playa depression it is essential that accu-mulation does not outweigh erosion (typicallydominated by deflation). Deflation is particularlyeffective, as the sediments are highly erodiblewhen exposed due to a combination of dis-persive agents (sodium), lack of protection fromvegetation and availability of fine sediments(section 5.2.2.2).

5.3.2 Ephemeral streams

Ephemeral systems (both sand and gravel bed)tend to act as temporary stores for sediment

Fig. 5.10 Death Valley playa with welldeveloped salt accumulation in theforeground. Note the irregular surface.The background consists of alluvial fansfeeding into the edge of the playa.Polygons are approximately 1 m across.

ESC05 9/9/06 2:42 PM Page 157

158 ANNE MATHER

within the arid environment, unless they crossor terminate in an area of relative tectonic sub-sidence. In addition, during periods of no flowthey may also act as a sediment source for aeolian processes. Ephemeral streams are domin-ated by large amounts of scour and fill withinsingle flood events (Leopold et al. 1966). Theperiod of subcritical flow during a flood event in ephemeral streams is short-lived (Reid &Frostick 1987). Thus the mechanism for the scouris also short lived and may relate to plane beds(Frostick & Reid 1987) and antidune migration(Foley 1978).

The bed scouring, together with abundantsediment supply, means that ephemeral riverscarry very high sediment loads (described as‘too thin to plough and too thick to drink’

by Colorado farmers; Beveridge & Culbertson1964). These flows can be extremely hyper-concentrated in sediment (Fig. 5.11). As theflood discharge increases, so does the sedimentload. Sediment supply exhaustion is unlikely inarid catchments, and ephemeral streams carry-ing 35 to 1700 times the sediment concentrationof perennial rivers are not uncommon (Frosticket al. 1983). In addition, suspended sedimentconcentrations tend to remain more in-phasewith flow characteristics than is observed inperennial rivers. Bedload transport in ephemeralrivers is also much greater than in perennialrivers. These differences may be 106 times largerat the threshold of entrainment and 101 at mod-erate shear-stress levels (Laronne & Reid 1993;Reid & Laronne 1995). In addition this trans-

1 10 100 1000 10,0001.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

99 91 80 50 40 20 9.1 5.0 0.99

Sediment concentration by weight (%)U

nit

wei

gh

t (t

m−3

)

Water content (%)

Hyperconcentrated Extreme High

Debris flow

Streamflow

Gran

ular p

asteF

luid

Debris flowEphemeral streamflowPerennial streamflowWet concrete

Fig. 5.11 The continuous spectrum of sediment concentrations from sediment-rich ephemeral rivers to debris flows. (Simplified fromHutchinson 1988.)

ESC05 9/9/06 2:42 PM Page 158


port is more predictable, probably reflecting thelack of armouring in arid-zone rivers, as the highsupply of slope sediment ensures that fines arenot preferably removed from the bed (Dietrichet al. 1989; Laronne et al. 1994). Large clasts(pebble to cobble) can be moved as much as 3 km in one flood event (Leopold et al. 1966).

General bedforms associated with rivers arediscussed in Chapter 3. Here the characteristicsmore commonly associated with the deposits of ephemeral rivers are highlighted. In theserivers tractional bedload tends to be dominatedby transverse and longitudinal mid-channel barforms. These comprise imbricated, sortedsediments associated with upstream dippinglow-angle beds representing the bar top. If waterdepth is sufficient an avalanche face and associ-ated cross-strata may develop. As a barform istypically associated with a coarser, upstream barhead and a finer downstream bar tail, migrationof the bar head over the bar tail will commonlylead to a coarsening upwards within the bardeposits. The finer part of the bedload andcoarser element of the suspended load tend to bedeposited as horizontal lamination in sand beds.These may develop from upper flow regime planebeds or from pulses of sediment-rich water super-imposed on the overall flood wave. The rapidlywaning flood stage and abundant fines (typicallymuds) that make up the extremely concentratedwater:sediment ratio tend to deposit thick (up to 10 cm) clay drapes. Once exposed and baked,these clay drapes have a tendency to desiccateand curl and be reworked in ensuing flows asmud clasts.

5.3.3 Alluvial fans

Alluvial fans are fan-shaped bodies dominatedby coarse sediment that has been transportedfrom steep upland catchments (Fig. 5.12). Theyrequire the juxtaposition of an upland sourcearea and a lowland sediment accumulation area,where the fan will develop. Their other mainrequirement is a source area capable of produc-ing course material, and high, flashy, peak dis-charges. These conditions are commonly met inarid environments and so fans form an integral

part of arid geomorphological systems (Harvey1997). Alluvial fans, as long as they remainuntrenched, tend to trap the deposits of eachsuccessive flood event originating from the associated fan catchment. Once alluvial fansbecome incised (trenched) throughout they willbegin supplying coarse sediment to adjacentenvironments such as playa and river systems.Thus untrenched fans provide a useful record of processes operative within mountain areasbordering basins in arid regions.

In general (i.e. in the absence of any majorexternal influences such as climate) alluvial fans tend to be dominated by debris flows in theyounger, more proximal parts of the alluvialfan. As the fan system matures, supplies of fine material become exhausted, the drainage networkexpands and the conditions for the generationof fluvial processes are increasingly met, leadingto the dominance of fluvial deposition in the fandispersal area. As the fluvial processes requirelower gradients than the debris flow processesto maintain transport, trenching of the proximalfan surface is often a common feature associatedwith mature fans (Harvey 1990).

5.3.3.1 Streamflow processes

Within alluvial fan accumulation areas the trans-port and deposition of sediment by streamflowmay occur in (i) clearly defined channels such as the fan head trench (Fig. 5.12), (ii) in wide, ill-defined channels or (iii) as unconfined flows(sheetfloods). The last of these is a relatively rarephenomenon (section 5.4.3). In arid regions,where sediment movement is typically transport-limited, sediments are deposited from flows thatare hyperconcentrated with sediment (40–70%by weight, Costa 1988) as well as more normalstreamflow events (less than 40% sediment byweight, Costa 1988; Fig. 5.11).

‘Normal’ streamflow leads to the depositionof imbricated, sorted sediments (see e.g. Nemec& Steel 1984; Costa 1988). Where sediment is ofa suitable calibre and water depths are sufficient(i.e. in confined areas of flow such as the fan-head trench), bars with an avalanche face maydevelop. Elsewhere, particularly in less confined

ESC05 9/9/06 2:42 PM Page 159

160 ANNE MATHER

areas of flow, sediment movement will be pre-dominantly as diffuse gravel sheets or longi-tudinal bars (Smith 1974). Fluvial processes willbe encouraged where the supply of fine material is limited, typically in larger catchments wherestream-side supply of sediment dominates.

Hyperconcentrated streamflows are commonon alluvial fans (Fig. 5.11). These flows mayhave some shear strength although there is some controversy as to their exact nature (Reid& Frostick 1994). Resulting sediments maycontain weak internal stratification, normal and

Fan apex

Watershed

Catchment(sediment production)Channel

(sediment transfer)

Alluvial fan(sediment accumulation)

Fan toe

(a)

(b)

Fig. 5.12 (a) Schematic representation of an alluvial fan system showing the sediment production (catchment), sediment transfer(channel) and sediment accumulation (fan) elements and (b) a view from the toe to apex of a typical alluvial fan in northern Chile. Notethe recent debris flow down the right-hand side of the fan.

ESC05 9/9/06 2:42 PM Page 160


reverse grading, weak imbrication and clastsmay be matrix-supported (Costa 1988).

Research has shown that fluvially dominatedfans may be restricted to certain lithologies suchas those associated with high-grade metamorphics(see for example the work by Harvey 1984, 1987),which source few clays and are dominated bycoarser material. The topographic constraints aretypically larger, lower relief catchments, whichcan produce higher water to sediment ratiosfrom the slopes. These types of drainage basinbecome increasingly abundant with the progres-sive erosional development of the catchment(i.e. as weathered slope material is progressivelyexhausted and larger, more open, lower reliefcatchments develop). The spatial extent of fluvialalluvial fans is typically greater than for thosedeposited by other processes as the potentialtransportation distance for the fluvial flows isgreater, but the gradient needed to maintaintransport is less. Thus fluvial-dominated fanstend to be larger and possess a gentler apex-to-toe gradient than those dominated by other flowprocesses (Harvey 1987, 1990).

5.3.3.2 Debris-flow processes

The transport and deposition of sediment bydebris-flow processes (Fig. 5.11) may be (i)confined within channels, or (ii) as an uncon-fined lobe. Debris flows behave as a plastic(Costa 1988) and may be channelized with levees in the upper parts of the fans, and moreunconfined and lobate in the more distal parts of fans. The types of flow will be determined bythe water to sediment ratio and the type of finesediment available. Essentially the flows can be(i) cohesive or (ii) non-cohesive, depending onthe shear strength of the flow resulting from thesediment concentration (typically 70–90% byvolume, Costa 1988) and nature of fines. Theresulting deposits of debris flows typically lacksorting, may possess inverse grading (towardsthe base) and normal grading (towards the top), are matrix-supported and lack much in theway of internal structure, but will vary in theirinternal organization depending on water con-tent and cohesivity of the flow (see Nemec &

Steel (1984), Postma (1986) and Costa (1988)for a discussion of mass flow processes). Debrisflows are associated with catchments that gener-ate sufficient fine material to lend shear strengthto the flow and are supplied predominantly byslope material, i.e. small, steep catchments (Wells& Harvey 1987). These conditions are typicallymet when the catchment is developed on sedi-mentary and low-grade metamorphic geologies(Harvey 1990), and before the drainage networkhas had time to reduce relief in the fan catch-ment. Debris flows require a steeper gradient tomaintain transport and tend to have a smallerrun out distance than comparable fluvial flows,so that debris-flow-dominated fans are typicallysmaller and steeper than corresponding fluvial-dominated fans.

5.3.3.3 Aeolian processes

Aeolian processes are not commonly associatedwith alluvial fans in the literature, yet observa-tions of fans in arid areas contradict this. It isnot uncommon to find impeded dunes (section5.3.4.2) developed against fan apexes, or devel-oped in topographic lows at the fan toe. Theaeolian material may subsequently be reworkedby debris and streamflow processes on the fansurface or buried by debris flows. Thus in someareas aeolian processes can be significant insourcing external fine material, and locallyaffecting flow processes on the fan.

5.3.4 Aeolian bedforms

Aeolian sand deposits in arid environmentscover approximately 5% of the Earth’s land sur-face and aeolian silt (loess) covers a further 10%.Dunes can vary in size from < 1 m to > 200 m in height and can represent the deposition of one windstorm event or, for larger systems, thedeposition of sediment over millennia. Dependenton source area, dunes are typically composed ofquartz sand, but can include any material capableof being wind blown (e.g. gypsum, volcanic ash,shell fragments). Where dunes are composed ofcarbonate or gypsum, cementation of the dunesmay occur early in their development.

ESC05 9/9/06 2:42 PM Page 161

Tab

le 5

.2Fr

ee, s

impl

e du

nes.

Not

e th

at su

rvey

s of d

une

type

s ind

icat

e th

at li

near

dun

es a

re th

e m

ost c

omm

on, f

ollo

wed

by:

tran

sver

se, p

arab

olic

(Tab

le 5

.3) a

nd st

ar d

unes

. Fea

ture

s suc

has

dom

e du

nes a

re c

ompa

rativ

ely

rare

(see

dat

a re

porte

d in

Fry

berg

er &

Gou

die

(198

1) a

nd T

hom

as (1

997b

)). F

or a

n ill

ustra

tion

of m

orph

olog

y se

e Fi

gs 5

.13

& 5

.16.

Cat

egor

y

Tra

nsve

rse

(con

sist

ent

unid

irec

tion

al w

inds

)

Lin

ear

(bi-

dire

ctio

nal

win

ds a

nd m

ore

vari

able

w

ind

dire

ctio

n)O

ther

(var

ious

w

ind

dire

ctio

ns)

Maj

or c

ontro

l on

form

Som

e di

rect

iona

l var

iabi

lity

in th

ew

ind

and

abun

dant

san

d su

pply

Les

s di

rect

iona

l var

iabi

lity

in th

ew

ind

and

abun

dant

san

d su

pply

Lim

ited

sed

imen

t sup

ply

A v

aria

ble

unid

irec

tion

al o

r bi

-dir

ecti

onal

win

ds, m

ore

limit

edsa

nd s

uppl

y th

an tr

ansv

erse

dun

esB

i-di

rect

iona

l win

ds

Opp

osin

g bi

-dir

ecti

onal

win

ds

Win

d di

rect

ions

may

var

yse

ason

ally

, or

whe

re tr

ansv

erse

dune

s m

igra

te in

to a

reg

ion

wit

hm

ulti

dire

ctio

nal w

inds

. San

d su

pply

is a

bund

ant a

nd w

ind

dire

ctio

nun

stab

leC

ompo

sed

of c

ours

e sa

nd, w

hich

limit

s m

ovem

ent t

o hi

ghes

tve

loci

ties

. Var

ious

win

d di

rect

ions

Var

ious

win

d di

rect

ions

Nat

ure

of m

ovem

ent

Forw

ard

mig

rato

ry

Ext

endi

ng(s

andp

assi

ngfo

rms)

Will

sho

w a

net

mig

rati

on if

one

win

d di

rect

ion

isdo

min

ant

Sand

acc

umul

atin

g,ve

rtic

al fo

rms

Lim

ited

Lim

ited

Inte

rnal

stru

ctur

e

Hig

h an

gle

cros

s-be

ds in

clin

ed d

ownw

ind

(mor

eta

bula

r se

ts in

tran

sver

se d

unes

and

a te

nden

cy to

wed

ge s

hape

d se

ts in

bar

chan

oid

dune

s). N

arro

wra

nge

of o

rien

tati

ons

of c

ross

-bed

s. C

ross

-bed

sfr

eque

ntly

trun

cate

d by

low

-ang

le e

rosi

on s

urfa

ces

that

dip

gen

tly

dow

nwin

dW

hen

view

ed in

cro

ss-s

ecti

on p

erpe

ndic

ular

to th

ew

ind,

set

s of

cro

ss-b

eds

are

seen

to d

ip in

opp

osit

edi

rect

ions

from

the

cres

tA

s ab

ove

but w

ith

mor

e co

mpl

ex b

eddi

ng p

atte

rndu

e to

the

sinu

ous

cres

tM

ix o

f cro

ss-b

ed o

rien

tati

ons

refle

ctin

g th

e bi

-dir

ecti

onal

mov

emen

t, a

ltho

ugh

one

dire

ctio

nm

ay d

omin

ate

Var

iabl

e an

d co

mpl

ex c

ross

-bed

s

?Wea

k

?Wea

k

Type

Tra

nsve

rse

ridg

eB

arch

anoi

dri

dge

Bar

chan

Lin

ear

ridg

e

Seif

Rev

ersi

ng

Star

Zib

ar

Dom

e

Slip

face

s

1 1–2

2 2 3+ 0 0

ESC05 9/9/06 2:42 PM Page 162


Dunes may occur as ‘simple’ dunes (where theindividual dunes create discrete forms), com-pound dunes (where simple dunes of the sametype coalesce and merge) or complex dunes(where different simple dunes merge). Here thefocus is on simple dunes, which, for the sake of simplicity, can be further subdivided into freedunes (Table 5.2 and Fig. 5.13), with a form dictated primarily by wind characteristics, andimpeded dunes (Table 5.3 and Fig. 5.14), a formthat is significantly affected by surface rough-ness or topographic barriers.

5.3.4.1 Free dunes

Free (or mobile) dunes are controlled primarilyby wind regime and sediment supply. Withoutavailable sediment or without wind to moveavailable sediment, dunes cannot form. The

nature of this relationship is illustrated in Fig.5.15, which plots wind direction variability(wind regime) against equivalent sand thickness(sand availability). Wind direction variability is measured by calculating the frequency andstrength of winds blowing from different direc-tions. A high value indicates a single predominant

Table 5.3 Impeded, simple dunes. For an illustration ofmorphology see Fig. 5.14.

Dune type Major control on form

Blowout Disrupted vegetationParabolic Vegetation anchoringLunetteShrub-coppice (Nebkha)Lee/fore dune Topographic barrierClimbing/falling duneEcho dune

(a) Transverse ridge

(c) Barchan

(d) Linear(e) Reversing

(g) Dome(f) Star

(b) Barchanoid ridge

Fig. 5.13 Morphology of the main free, simple dunesdescribed in Table 5.2. (Adapted from McKee 1979.)

ESC05 9/9/06 2:42 PM Page 163

164 ANNE MATHER

wind. A low value indicates a highly variable windregime. Equivalent sand thickness represents thethickness of dune sand in an area if it were spreadevenly, so the greater the thickness, the greaterthe overall volume of the dune sand. Figure 5.15

shows that the largest dunes on Earth, star dunes,form in areas with greatest wind variability andgreatest sediment supply. The calibre of availablesediment is also important in the generation ofdune types. For example, coarse sand is usuallyassociated with dunes with low-angle morphologyand limited height (typically up to 4 m). Thesedunes may be linear (zibar dunes) or mound-like(dome dunes) depending on wind regime. Strongwind velocities are required for transport, andmovement of the dunes is limited.

Initial dune-sand accumulation may be initi-ated by a surface roughness or by temperaturegradients in the atmosphere (sand accumulationtends to occur where airflows are ascending andrelatively slower). The accumulation and move-ment of sand will eventually develop a stoss-and slip-face. Where there is abundant supply of suitable sediment and unidirectional wind,transverse ridge dunes will develop (Fig. 5.15).If two opposing prevailing winds of similarstrength and duration exist, reversing dunes maydevelop as active slip faces change with winddirection. As topographically lower dunes canmove more quickly than higher dunes, a sinuos-ity develops, creating a sinuous, barchanoid ridge

(a) Blowout (b) Parabolic dune

(c) Lunette (d) Nebkha

(e) Lee dune (f) Climbing dune

EchoduneFore

dune

Fallingdune

Fig. 5.14 Morphology of the main impeded, simple dunes described in Table 5.3. (Adapted fromSummerfield 1991.)

00

10

20

30

Equ

ival

ent

sand

thi

ckne

ss (

m) 40

50

0.2 0.4Wind direction variability

Multidirectional Unidirectional

0.6 0.8 1

Star

Transverse

LinearBarchan

Fig. 5.15 Relationship between wind variability and equivalentsand thickness (thickness of sand if it were spread evenly over anarea). For explanation see text. (After Wasson & Hyde 1983.)

ESC05 9/9/06 2:42 PM Page 164


(Fig. 5.16b). Eventually this pattern will give riseto a dune network of nearly linked ridges trend-ing at 10–20° to the prevailing wind direction(see Warren (1979) for more details on airflowover dunes).

In contrast, under similar unidirectional windconditions to those that form transverse andbarchanoid ridges, but where sand supply is lessabundant, barchan dunes will form (Figs 5.15& 5.16a). The lower outer ridges will movemore quickly than the higher, central portions,leading to significant arms that extend down-wind. The width and spacing of the dunes willreflect the effects of airflow patterns, which maybe related to upwind dune forms.

In areas of high wind variability, but relativelylimited sediment supply, linear dunes are created(Fig. 5.15). Linear dunes (also called longitudinaldunes, and including sinuous linear dunes or seifdunes) are by far the most common dunes onEarth. They are highly elongated in form and con-tain two opposing slip faces either side of a crestline. Linear dunes can be kilometres in lengthand coalesce downwind into Y-shaped junctions.They are typically associated with obliquely con-verging wind directions, although the details oftheir formation are still debated (see discussionsin Thomas 1997b).

Star dunes can be 300–400 m in height andcontain the largest sand volume (Figs 5.15 &5.16c). This dune form requires abundant sediment supply and multiple dominant winddirections (Fig. 5.15). The latter wind directionsare typically associated with seasonal variations,which cause other dunes to merge and modify.Thus star dunes are common in the centre ofsand seas, although they are also associatedwith topographic barriers that influence regionalwindflow (Lancaster 1994). Once sufficientlylarge, star dunes can modify their own windflowpatterns.

5.3.4.2 Impeded dunes

Impeded (or anchored) dunes are dominated by topographic barriers that inhibit sand move-ment and lead to sand build-up. The variety offorms is illustrated in Table 5.3 and Fig. 5.14.

The most common type of these dunes is theparabolic dune. Parabolic dunes tend to developwhere blowouts occur. These are formed by the deflation of sand (often where vegetationhas been disturbed), which leads to areas ofhigher wind speed over the associated lowerroughness surfaces. The increased wind speedfurther dries the sand making it more prone to mobilization. With increased deflation theblowout will increase in size and the leewardrim will migrate downwind, leaving trailinglimbs which are ‘anchored’ by vegetation. Thesetend to limit the main sand movement in thehigher, central part.

5.3.4.3 Internal structures of dunes

Dunes form through the accretion of tractionaldeposits or a drop in wind velocities (and thusreduction in shear stress) leading to grain-falldeposition. Once the initial sand has becometrapped, sediment will continue up the stoss(windward) slope to the crest of the dune. Herematerial will avalanche down a slip face. Thusmost dune types are associated with particularsedimentological characteristics (Table 5.2). Theinternal structure of sand dunes is dealt with inmore detail in McKee (1979). This and otherresearch is summarized in Nickling (1994).

Free dunes (which dominate in arid to hyper-arid climates) are typically associated with welldeveloped cross-strata which dip downwind at30–34°. Dune deposits are also associated withoften massive, tabular to planar cross-strata thatthin from the base upward. Another commonfeature is the development of bounding surfaceswhich occur between sets of cross-strata, whichresult from the migration of dunes over inter-dune areas.

Vegetated dunes (which are typically associatedwith semi-arid climates) tend to show a bimodaldistribution of steeper angle cross-strata andminor truncation surfaces. Most dips are low(about 12°). Parabolic dunes are dominated bysteeper cross-strata (similar to transverse dunes),which often have concave slip-faces. They mayalso contain some concave-upward sets that havebeen deposited in hollows. Vegetated dunes are

ESC05 9/9/06 2:42 PM Page 165

166 ANNE MATHER

(a)

(b)

(c)

Fig. 5.16 Dune types sourced from wadi sand in the Draa river, southern Morocco: (a) barchan dunes, (b) barchanoid ridge (50 m from stoss to lee) and (c) star dune (300 m across – arm to arm).

ESC05 9/9/06 2:43 PM Page 166


also associated with evidence of vegetation suchas rhizoliths.

5.3.5 In situ landform modification

Once sediment has been deposited, depending onits residence time, it can undergo some form ofin situ modification through pedogenic, biogenicor hydrological processes. These processes varyspatially in their intensity, creating a mosaic ofsurface runoff sources and sinks during a rain-storm. This is significant for sediment entrain-ment, erosion and deposition as is illustrated in section 5.4.3. The most significant in situmodifications of the sediment surface are dis-cussed in sections 5.3.5.1 to 4 below.

5.3.5.1 Arid region soils

The USA soil taxonomy classification system re-cognizes five main soil orders (Table 5.4) for aridregions. It is evident from this table that arid areasare dominated by poor soils lacking in organicmatter and associated with limited plant growth.Arid soils differ from soil developed in morehumid regions in a number of ways. Typicallyaccumulation is dominated by material fromexternal sources (e.g. aeolian dust), rather thanthe in situ breakdown of parent material obser-ved in more humid environments. In additiondesert soils are typified by long time-scales ofdevelopment. This facilitates the long-term buildup of minerals such as carbonate, or gypsum in hyper-arid settings (Fig. 5.17a), developingindurated ‘petrocalcic’ horizons (section 5.3.5.5).

5.3.5.2 Desert pavements

‘Desert’ or ‘stone’ pavements are typified by alag of stones, typically overlying a stone-free silt(Fig. 5.17b). Where developed the stones act toprotect the surface from erosion, by breaking up runoff flow paths. It is widely believed nowthat the fines are largely aeolian in origin andconcentrate around the stone lag, passing intothe regolith, perhaps along desiccation cracks,displacing the stones by pedogenic and aeolianaccretion (McFadden et al. 1984, 1987). Thestones are too large to follow and remain at thesurface. Some authors (e.g. McFadden et al. 1987)have argued that the enrichment with clay andcarbonate restricts water penetration into the landsurface in the initial stages of surface develop-ment, shedding more runoff and instigating gullyincision, which will eventually isolate the sur-face, deactivating soil processes and facilitatingfurther pavement development. Any associatedcoarse material at the surface can also affect thesurface hydrology. Lavee & Poesen (1991) foundthrough simulated experiments that in generalincreased stone cover will enhance runoff gener-ation. The response, however, is complex, as the stone cover can also act as a mulch, trappingmoisture and encouraging plant growth. In suchsituations vegetation may inhibit surface sealdevelopment and encourage infiltration.

5.3.5.3 Microphytic crusts

Although desert soils are typified by low vascular plant cover, they can have abundant

Table 5.4 Main soil orders associated with arid environments in order of abundance. (Adapted from Dregne 1976; Dunkerley & Brown1997.)

Soil order Main characteristics Total land occupied (106 km2)

Entisol Minimal horizon differentiation and modification of the bedrock 19.2 (sedimentary material)

Aridisol Dryness and/or salinity restricts plant growth throughout the year 16.6Mollisol Thick, dark base-rich and organically rich topsoil or surface horizon (epipedon) 5.5Alfisol Moderate base saturation, argillic horizon and seasonally available water for 3.1

plant growthVertisol Deep clay soils, characterized by cracking and shrink and swell characteristics 1.9

ESC05 9/9/06 2:43 PM Page 167

168 ANNE MATHER

non-vascular cover (e.g. mosses, algae, lichens,fungi, bacteria – microphytes). Where micro-phytic crusts are developed they are typicallyirregular in topography and act to break up run-off on the surface, protecting it from rill erosion.

In some environments these crusts act as perme-ability barriers, and increase runoff, but protectthe underlying surface. However, the impacts ofthe crusts are widely variable (see the discussionin Dunkerley & Brown 1997).

(a)

(b)Fig. 5.17 In situ modification ofsediments in arid environments: (a) gypcrete development and (b)desert pavement developed in theAtacama Desert of Northern Chile. In (b) note the uniformity of the mainsurface clast size and dominance of silt grade material beneath thesurface in the linear surface scrape(light colour) to the left of the pencil.Note also the discoloured (red oxide,appears as light grey in the image)base of the upturned clast lyingunderneath the pencil. The darkishcolouration of most of the surfaceclasts is due to the development ofdesert varnish (see Oberlander(1994) for a summary of how thisforms). Hammer is 40 cm long andpencil is 15 cm long.

ESC05 9/9/06 2:43 PM Page 168


5.3.5.4 Inorganic (soil or rainbeat) crusts

Inorganic crusts may develop on soil surfacesduring or shortly after a rainstorm event. Theserelate to physico-chemical changes relating towater and clay content and rainfall characteristics.These can be grouped into (i) structural seals or (ii) depositional seals (Maulem et al. 1990).Structural seals develop as a result of the soilstructure as a function of aggregate breakdownby rainsplash and slaking; washing of fines intoadjacent, larger pores; particle segregation andcompaction. Depositional seals are a function of the deposition of fine suspended sediments ina thin, comparatively impermeable film on thesoil surface. Sealing intensity has been demon-strated to decrease with increasing slope angleas a function of higher erosion on higher slopes(Poesen 1986). Poesen also found that embeddedrock fragments in sandy soils can generate seals,whereas those with rock fragments sitting on thetop do not. This was attributed to the fact thatnon-embedded stone particles typically have‘overhangs’ that protect the area immediatelyadjacent to the rock fragment from raindrop im-pact, thus a complete soil seal cannot be created.Once developed, surface seals may prevent theescape of air as a wetting front penetrates the soil,generating associated vesicular layers 1–30 mmbelow the surface (Chatres et al. 1985; Ringrose-Voase et al. 1989). These vesicles further reducehydraulic conductivity of the soils.

5.3.5.5 Duricrusts

These are indurated surface/near-surface crusts.They may be exhumed relict deposits (e.g.ancient lacustrine deposits) or may relate tocontemporaneous hydrological or pedologicalprocesses in the desert environment. Some crustsare ephemeral, for example salcretes (Table 5.5),others may persist for many millions of years(such as the gypcretes reported by Hartley &May 1998). Table 5.5 describes the duricrustsmost commonly found in arid environments.Note that there are many more subtypes, such as the nitrate deposits in the Atacama Desert ofnorthern Chile.

5.4 NATURAL IMPACTS ON PROCESSES

Arid environments, as described previously, are dynamic. Their spatial extent and degree ofaridity can change over a range of temporal andspatial scales (Fig. 5.4). Although some of thesechanges can be accounted for by intrinsic con-trols, extrinsic controls can play an importantrole. These consist of external forcing mechan-isms that cause the arid system to change. Whatis considered to be the dominant external con-trol will change depending on the scale of thelandscape element examined and the time-scaleinvolved. For example, long term (million year)time-scales are typically dominated by tectoniccontrols due to the large spatial and temporalscale on which they cumulatively operate. Climatecontrols tend to dominate the medium scale(million to thousand year) and short term time-scales (annual) are dominated by factors such asstorminess.

5.4.1 Long term time-scale controls: tectonics

Active tectonics can have an impact on arid envir-onment sediment systems directly or indirectly.Direct effects may have an impact on localizedareas of an arid landscape, for example theymay generate fault scarps as a function of anearthquake event, which can then divert localdrainages, or rejuvenate local erosion. These typesof features, however, are typically ephemeral, andare removed in a matter of years, depending onthe rates of erosion. Even within hyper-arid areaswhere water availability is low, in situ weather-ing, aeolian deposition and gravitational col-lapse can conceal the effects of faulting relativelyquickly. Within alluvial fan systems tectonics candirectly rejuvenate the catchment areas that pro-duce and deliver the sediment to the alluvial fan,for example by increasing slope angles throughuplift. Within the sediment accumulation areaof the alluvial fan itself tectonics may locallydirectly modify the fan surface. The morphologyand internal sedimentology of alluvial fans canbe seen at its simplest as a balance between thedischarge of sediment exiting the fan source areaand the available accommodation space. Thus

ESC05 9/9/06 2:43 PM Page 169

Tab

le 5

.5Th

e m

ajor

type

s of d

uric

rust

foun

d in

arid

env

ironm

ents

. Sol

ubili

ty in

crea

ses f

rom

silc

rete

to sa

lcre

te. (

Sum

mar

ized

from

Wat

son

& N

ash

1997

and

refe

renc

es th

erei

n.)

Cru

st ty

pe

Silc

rete

Cal

cret

e

Gyp

cret

e

Salc

rete

Mai

nm

iner

al

Silic

a

Cal

cium

carb

onat

e

Gyp

sum

Hal

ite

Fiel

d ap

pear

ance

Bri

ttle

, int

ense

ly in

dura

ted.

Col

our

is v

aria

ble

wit

h gr

ey b

row

n an

d gr

een

com

mon

.

Occ

urs

as p

owde

ry th

roug

h to

nod

ular

,la

min

ated

to h

ighl

y in

dura

ted.

Gen

eral

lyw

hite

, cre

am o

r gr

ey w

ith

pink

mot

tlin

g an

dba

ndin

g.

Occ

urs

as 1

) hor

izon

tally

bed

ded

crus

ts;

2) s

ubsu

rfac

e cr

usts

wit

h la

rge

(1m

m –

0.

5m

dia

met

er) c

ryst

als

(des

ert r

oses

);

3) m

esoc

ryst

allin

e (c

ryst

als

50μm

–1

mm

) or

4) s

urfa

ce c

rust

s of

ala

bast

rine

gyp

sum

(cys

talli

tes

<50

μm).

Occ

ur a

s co

lum

nar

crus

ts, p

owde

ry d

epos

its

or s

uper

ficia

lco

bble

s. C

olou

r ra

nges

from

whi

te/g

rey

togr

een

or r

ed.

Typ

ical

ly w

hite

, but

can

be

pink

whe

re

mic

ro-o

rgan

ism

s ar

e pr

esen

t.

Typi

cal p

rofil

eth

ickn

ess

0.5–

3m

1–5

m

0.1–

5m

0.1–

5m

Mod

es o

f orig

in in

arid

envi

ronm

ents

Pedo

geni

c an

dgr

ound

wat

er.

Hig

hly

alka

line

evap

orit

ic b

asin

s or

hydr

olog

ical

or

pedo

geni

c or

igin

s.

Pedo

geni

c,gr

ound

wat

er a

ndla

cust

rine

eva

pori

tes.

Eva

pora

tion

of w

ater

sin

Sab

kha

sett

ings

,gr

ound

wat

er a

ndpe

doge

nic

(the

latt

ertw

o be

ing

rare

r).

Dis

tribu

tion

Foun

d in

are

as n

ot s

ubje

cted

to la

te T

erti

ary

and

Qua

tern

ary

glac

iati

on. L

ess

glob

ally

sig

nific

ant t

han

calc

rete

. Man

tle

low

gra

dien

t slo

pes,

but

can

als

ofo

rm b

enea

th b

asal

tic

lava

flow

s an

d in

wea

ther

ing

profi

les

on a

ran

ge o

f lit

holo

gies

.M

ost w

ides

prea

d cr

ust.

Pet

roca

lcic

hor

izon

s co

ver

20×

106

km2

of th

e gl

obe’

s la

ndsu

rfac

e. T

ypic

al o

far

eas

whe

re m

oist

ure

is d

efici

ent t

hrou

ghou

t the

seas

ons.

Ped

ogen

ic c

alcr

etes

man

tle

undu

lati

ng o

rge

ntly

slo

ping

terr

ain.

Can

form

on

mos

tlit

holo

gies

.T

ypic

al o

f war

m d

eser

ts w

ith

very

low

rai

nfal

l (<

250

mm

a−1).

Ped

ogen

ic g

ypcr

etes

man

tle

low

grad

ient

sur

face

s. T

hey

are

also

ass

ocia

ted

wit

hhy

drol

ogic

al b

asin

s. C

an d

evel

op o

n a

rang

e of

litho

logi

es.

Ass

ocia

ted

wit

h ar

eas

of e

xtre

me

arid

ity,

part

icul

arly

in a

ssoc

iati

on w

ith

salin

e ba

sins

and

rain

falls

less

than

200

mm

a−1. C

an fo

rm th

inep

hem

eral

cru

sts

on d

une

sand

s.

ESC05 9/9/06 2:43 PM Page 170


spatially variable subsidence in the fan accumu-lation area can go some way to controlling fanthickness and morphology (Whipple & Trayler1996). Viseras et al. (2003) found that in theBetic Cordillera of southern Spain in areas wherethe accommodation space for alluvial fan devel-opment is driven by relatively high tectonic sub-sidence (at > 1 mm yr−1) the resultant alluvial fanstended to aggrade vertically and were not incised.Where the accommodation space was controlledby lower rates of subsidence (< 1 mm yr−1), how-ever, the resultant fans were narrower and pro-graded out into the basin with some incision.

Indirect effects of tectonics include situationswhere regional tectonics, such as epeirogenicuplift, create the necessary conditions for a secondary geomorphological process to operate

such as river capture. This situation tends to occurwhere the development of regional gradients canincrease the stream power, and thus erosivity, of one stream in relation to its neighbours. Theheadward erosion associated with these pheno-mena may be exacerbated where the geologicalstructure generated by the tectonics also exposessediments of higher erodibility. Combined withthe runoff-dominated hydrology, high-intensityrainstorm events and minimal protection fromvegetation cover associated with arid envir-onments, and river capture becomes a prolificmodifier of drainage networks. As such rivercapture can play a major role in sediment pro-duction, delivery and routing in tectonicallyactive arid landscapes. Details of such an exam-ple are highlighted in Case Study 5.1.

Case study 5.1 Influence of tectonics on rates of sediment production, delivery and routing, the Sorbas Basin, south-east Spain

The Sorbas Basin is located within a basin and range topography within the Betic Cordillera of southern Spain. Southern Spain represents part of the plate boundary between the Africanand Iberian plates, which ceased subduction in the late Neogene. Since this time compressionhas dominated with tectonic movement being expressed through differential uplift within andbetween the sedimentary basins and the mountain ranges. Pliocene to recent average uplift ratesare calculated to be in excess of 160 m Myr−1 for the Sierras Alhamilla and Cabrera, but aretypically much lower for the basins (80 m Myr−1 for the Sorbas Basin; 11–21 m Myr−1 for theVera Basin; Mather et al. 2002). This deformation has been significant in generating regionaltopographic gradients. Combined with the geological structure this has led to the develop-ment of abundant river captures as a direct function of related increases in stream power andaccelerated headward erosion. These river captures are significant in affecting the sedimentflux, routing and delivery within and between basins (Mather et al. 2002; Stokes et al. 2002).One such capture occurred in the upper Pleistocene and forms the focus of this case study.

The modern topography of the Sorbas Basin is dominated by an incising drainage network(the Rio Aguas), which is associated with many landscape instabilities such as extensive badlandterrain and landslides (Case Fig. 5.1a). Landscape erosion is thus locally severe (Harvey 2001),with some areas of abandoned agricultural land undergoing gully headcut retreat of several metresin one rainstorm event. Erosion-pin experiments indicate localized surface lowering of severalmillimetres per year. In some lithologies piping is abundant and forms a pseudokarst system,which supplements and overprints surface runoff features in terms of sediment and water discharge and routing. Localized piping features may be as much as 15 m deep and 1–2 m indiameter. These badland areas are associated with material of low residual strength such as marlsand silts. The main landslides, in contrast, tend to dominate in areas of stronger lithologies withhigher unconfined compressive strengths such as limestones, or in pervious material, which in

ESC05 9/9/06 2:43 PM Page 171

172 ANNE MATHER

Case Fig. 5.1 (a) Distribution of erosional and landslide features in relation to the 100 ka river-capture site. Note the increasein frequency and magnitude of landslide occurrences in proximity to the river-capture site, and greater abundance of badlandareas. The box outlines the area depicted in (b). (From Griffiths et al. 2002.) (b) Surface lowering above river-capture site. Notethat the largest amounts of surface lowering are associated with (i) the valley networks, (ii) valley confluences and overall with(iii) the lower reaches of the drainage network proximal to the capture point. S, Sorbas. (Adapted from Mather et al. 2002.)

2 km

NRambla de Sorbas Río Aguasde

Ram

bla de los

Feos

Rambla de los

Chopos

Rambla deCinta Blanca

River

Watershed, Río de Aguas

Watershed, Rambla de los Feos

X Badlands

Landslides

>10 m5 3

>10 −10 m3 5 3

<103 m3

X Capture site

73 74 75 76 7778 79

13

12

11

10

09

08

07

06

05

80 81

100

(b)

(a)

80

60

40

20

0 (m)

Modern drainage network

Surface lowering

Rbla de Cinta Blanca

Rbla de Góchar

Rbla de S

orbas/

Upper Aguas

SRbla de

los Chopos

Rambla de

Góchar

ESC05 9/9/06 2:43 PM Page 172


an arid environment has more resistance to erosion, such as gypsum. The landslide featuresvary in style from deep-seated rotational slips to block topples and may be up to 1 million cubicmetres in volume (Griffiths et al. 2002).

The spatial and temporal distribution of the above features in the landscape is governed by their geographical location to the upper Pleistocene river capture site. The beheaded river(Rambla de los Feos in Case Fig. 5.1a) experienced a reduction in erosion rates. In contrast thecaptured drainage system (Upper Aguas; Ramblas Góchar, Cinta Blanca and Chopos in CaseFig. 5.1a & b), and the area below the capture site (Lower Aguas), incurred dramatic increasesin net erosion. At the capture site base level was lowered by 90 m, starting the propagation of a rejuvenating wave of incision up the catchments and accelerating sediment production anddelivery to the fluvial system (Mather et al. 2002; Stokes et al. 2002). The river capture eventwas thought to have occurred at around 100 ka. More recent U-series dating, however, sug-gests that it may be much younger, before 77.7 ± 4.4 ka (Candy et al. 2005). Over this timeframe the impact of the river capture has reached 20 km upstream, at a decaying rate. Near thecapture site a tenfold increase in incision was experienced. This changed the sediment deliveryprocesses. Initially the generation of steep, rapidly unloaded slopes generated mass movementprocesses such as landslide failures (Mather et al. 2003). In weaker lithologies the dominantsediment delivery process later became dominated by more progressive slope erosion by surfacerunoff and subsurface piping processes (pseudokarst). Most of this accelerated erosion is stillrestricted to the main valley side-slopes, and has not yet reached the main drainage divides sothat overall surface lowering is much less than that recorded by the localized valley incision(Case Fig. 5.1b).

The change in processes driven by the tectonically induced river capture has affected thePleistocene to Holocene sedimentation within the study area in a number of ways. In the pre-capture state the then linked Upper Aguas and Feos fluvial terraces were dominated by largematerial (pebbles, cobbles and boulders) sourced from the Sorbas Basin and routed to the sedi-mentary basins to the south. Post-capture, however, the beheaded section of the river system(the Feos) underwent a reduction in sediment and water discharge and became dominated bylocalized and smaller scale slope erosion. This is reflected in a change in sedimentology of thefluvial terrace deposits in terms of sediment calibre, provenance and sedimentary style. Post-capture fluvial terrace deposits are dominated by more localized, typically smaller (granule andpebble) material than the pre-capture terraces and no longer receive sediment sourced fromwithin the Sorbas Basin. In addition the pre-capture river terrace deposits are frequently buried by fine-grained colluvium in the more open parts of the valley and coarser, alluvial-fanmaterial in the more valley constrained, basement sections of the river. In contrast the capturedUpper Aguas and Lower Aguas have seen enhanced sediment production and delivery to the fluvial system, together with the re-routing of the high water and sediment dischargessourced from the Sorbas Basin to the sedimentary basins to the east. Adjacent to the capture sitethe sediment delivery from the valley sides has been dominated by a combination of landslidingand gullying.

Relevant reading

Candy, I., Black, S. & Sellwood, B.W. (2005) U-series isochron dating of immature and mature calcretes as abasis for constructing Quaternary landform chronologies for the Sorbas Basin, southeast Spain. QuaternaryResearch 64, 100–11.

ESC05 9/9/06 2:43 PM Page 173

174 ANNE MATHER

5.4.2 Medium term time-scale controls: climate

Climate and humans can have very similareffects on landscapes vulnerable to change, suchas those in the arid realm, as they both affectvegetation cover (style and abundance) andthrough this have an impact on the erosional,transportational and depositional elements of

the arid system (Case Study 5.2). Typically the older the record then the less likely humanimpact is to be the main control. This reflectsboth the existence and abundance of people insufficient numbers to drive change and also the technology available to implement change.Thus on Pleistocene time-scales climate is morelikely to be the driving force.

Case study 5.2 Climate controls on sediment production and delivery, the pluvial lakes of the Basinand Range Country, American South-west

Many of the basins in the regions of Nevada and California were once the home of extensivelake systems. The high stands for these lake systems mainly correspond with Quaternary glacial periods in the high Sierra Nevada when the adjacent lowland areas (basins) were sub-jected to relatively wetter and cooler climate conditions. The lowest of these basins was (and is)Death Valley. There is some debate over the exact nature of connectivity between drainages,but it is suggested that possibly three rivers fed Lake Manly of Death Valley (Jannik et al.1991). These were (i) the Amargosa (rising from the Spring Mountains), (ii) the Mojave River(San Bernadino Mountains) and (iii) Owens River (Sierra Nevada). This water passed throughfour pluvial lakes: Owens, China, Searles and Panamint (Case Fig. 5.2A). In some cases theoverflow points between the lakes are clear, for example fossil falls (Case Fig. 5.2B) betweenOwens Lake and China Lake. In others, such as the link between Lakes Panamint and Manly,the evidence is less clear. Evidence of former lake levels comes from shorelines marked bybenches, shoreline tufas and beach bar deposits (Case Fig. 5.2C).

Using dated shorelines as a framework it has been established that lakes fluctuated in level temporally (Jannik et al. 1991). It also can be demonstrated that for the same time periodsediment supply changed spatially. Harvey et al. (1999) demonstrate that the Lakes Lahontanand Mojave (Case Fig. 5.2A) show quite different sedimentary responses to the same late

Griffiths, J.S., Mather, A.E. & Hart, A.B. (2002) Landslide susceptibility in the Rio Aguas catchment, SE Spain.Quarterly Journal of Engineering Geology and Hydrogeology 35, 9–17.

Harvey, A.M. (2001) Uplift, dissection and landform evolution: the Quaternary. In: A Fieldguide to theNeogene Sedimentary Basins of the Almería Province, SE Spain (Eds A.E. Mather, J.M. Martin, A.M. Harvey& J.C. Braga), pp. 225–322. Blackwell Scientific, Oxford.

Harvey, A.M. & Wells, S.G. (1987) Response of Quaternary fluvial systems to differential epeirogenic uplift:Aguas and Feos River systems, south-east Spain. Geology 15, 689–93.

Mather, A.E., Stokes, M. & Griffiths, J.S. (2002) Quaternary landscape evolution: a framework for understand-ing contemporary erosion, SE Spain. Land Degradation and Management 13, 1–21.

Mather, A.E, Griffiths, J.S. & Stokes, M. (2003) Anatomy of a ‘fossil’ landslide from the Pleistocene of SE Spain.Geomorphology 50, 135–49.

Stokes, M., Mather, A.E. & Harvey, A.M. (2002) Quantification of river capture induced base-level changesand landscape development, Sorbas Basin, SE Spain. In: Sediment Flux to Basins: Causes, Controls and Con-sequences (Eds S.J. Jones & L.E. Frostick), pp. 23–35. Special Publication 191, Geological Society PublishingHouse, Bath.

ESC05 9/9/06 2:43 PM Page 174


Pleistocene to Holocene climatic transition. The alluvial fan systems of Lahontan, sourced fromthe Stillwater Range underwent minimal slope erosion and fan sedimentation during that timeframe. In contrast the fan systems of the Mojave, sourced from the Transverse Ranges of theSan Bernadino Mountains, were far more active over that same period, first dominated by debrisflows on the fan surfaces and then by ephemeral streamflow causing fanhead trenching and

Case Fig. 5.2A Pleistocene lakes of the American South-west (Adapted from Jannik et al. 1991; Harvey et al. 1999.)

O r e g o n I d a h o

N e v a d a

C a l i f o r n i a

P a c i f i cO c e a n

Arizona

Los Angeles

Las Vegas

Soda Lake

Silver Lake

HarperLake

CoyoteLake

TecopaLake

SearlesLake

Panamint LakeOwensLake

Mono Lake

Lake Tahoe

Pyramid Lake

LakeLahontan

Lake Mojave

Lake Mannix

Lake Manly

Long Valley Lake

Abobe Lake

ChinaLake

PahrumpLake

Reno

Transverse Ranges

(San Bernadino Mts)

Si

er

ra

Ne

va

da

SODAMTS

Ra

ng

e

Sti

llwa

ter

Ow

ens

River

Car

so

n R.

Wal

ker

Rive

r

Mojave River

Colorado

River

HumboltR

iver

0 miles 100

0 km 150

? ?

Troy Lake

Death

Valley

Am

argosa

River

Present day lakes & playas

Pleistocene lakes

Present day rivers

Pleistocene rivers

Catchment boundary

State boundary

N

ESC05 9/9/06 2:43 PM Page 175

176 ANNE MATHER

Case Fig. 5.2B Fossil falls overflow between Owensand China lakes.

Case Fig. 5.2C (below) Former shorelines of LakeManly, Death Valley. (a) Tufa lines (arrowed) atBadwater and (b) the Beach Ridge on the road to Beatty.

(a)

ESC05 9/9/06 2:43 PM Page 176


distal alluvial fan sedimentation and progradation. Combining this information on vegetationcover with climate reconstructions (Case Fig. 5.2D) highlights the interaction between thesetwo variables and their impact on sediment production and delivery. In the Late Pleistocene bothLahontan and Mojave were subjected to much wetter and cooler climates than today. Winterprecipitation was dominant and heavy snowfall in the Sierras and the Transverse ranges main-tained the pluvial lake levels. During the Late Pleistocene to Holocene transition the weakeningof zonal circulation and the monsoonal incursion of warm moist tropical air from the Gulf ofMexico or eastern Pacific affected the Lake Mojave catchment. This meant that summer stormsbecame more effective in this region. The Lahontan area was not affected by these storms. Interms of the vegetation cover, research using the distribution of remains from packrat middensdemonstrates that catchment hillslopes in the Mojave supported a desert shrub vegetation. Incontrast the Lahontan supported juniper woodland and grasses at low elevations and pine at

Case Fig. 5.2C (Continued)

(b)

Case Fig. 5.2D Postulated Late Pleistocene to Holocene changes in airmass circulation. (From Harvey et al. 1999.)

H

H

Easterlies

Jetstreamto the South.Rain andsnow to theSierras,TransverseRanges, etc.

Meridional flowallows summer

monsoonal penetrationof tropical air in the

southwest (but rarely into the Great Basin)

Jetstreamshifts north.Winter rainin NorthernSierras

0 500

km

18 ka (Last Glacial Maximum) 14 –11 ka 11–8 ka

Late Pleistocene - Early Holocene conditions

Glacial ice(approximate)

Zones of heavyprecipitation

High pressure

Zones ofheavy (winter)precipitationH

Lake Lahontan(rel. high)

Lake Mojave(high)

Cool moist

Cold dry

Cool wet

Strong zonalcirculation

Cool moist

Lake Mojave(high)

Lake Mojave(falling)

Winterprecipitaion

Convectional(summer)rainfall

Low pressurecentres

L

Jetstream wellto the South.Rain and snow inWestern Sierras,Transverse Ranges, etc. Great Basin, cold and dry.

L

Lake Lahontan(falling)

Lake Lahontan(max)

ESC05 9/9/06 2:43 PM Page 177

178 ANNE MATHER

Alluvial fans may be affected by changes inclimate, where changes in process are affected.For example, where climate either limits orenhances the weathering processes within thecatchment the water to sediment ratio of thedominant fan process may change. Similarly,changes in the amount of effective runoff gener-ated may again change the water to sedimentratio and thus modes of sediment transport and accumulation (Case Study 5.2). Streamflowprocesses require less of a gradient for trans-port than debris-flow fans, so a switch fromdebris flow to streamflow may lead to fan-headtrenching. If the entire fan becomes trenched,the mountain supply catchment will then beable to supply sediment to the main basin area,altering basin sediment routing. Alternativelyclimate may control lake-levels in arid environ-ments (Case Study 5.2). As with tectonic con-trols, lake-level fluctuations can control alluvialfan morphology and sedimentary architectureby controlling the accommodation space foralluvial fan development. Lake-level rise willtypically lead to retrogradation of the alluvial

fan system, backfilling into mountain embay-ments. Viseras et al. (2003) demonstrated thatthis is associated with a reduction in the size offeeder channel. In these cases the fan morpho-logy most strongly reflects the morphology ofthe mountain embayment in which the fan wasdeveloped, rather than the nature of the catch-ment area. This reflects the control that theembayment has on the distribution of debrisand ephemeral streamflows over the alluvial fansurface (Viseras et al. 2003).

Arid aeolian systems are particularly sensitiveto changes in precipitation and also to factorssuch as temperature, affecting evapotranspira-tion. The threshold of change for precipitationlies within the arid–semi-arid transition zone,where increased vegetation cover becomes moreimportant. Thus if precipitation increases dunetypes may revert from free to vegetationallyimpeded (section 5.3.4) and existing dunes maybecome stabilized. In addition, aeolian envir-onments will be susceptible to changes boughtabout by changes in wind strength and direction.Examination of Quaternary records in tropical

higher elevations. Thus in Lahontan the more continuous vegetation cover reduced the generationof runoff and sediment production, leading to only rare debris-flow events associated with thewinter storms. In contrast the sparse, discontinuous vegetation of Mojave allowed greater runoffgeneration from the winter rainfall, and associated enhanced overall sediment erosion and deliveryin the form of debris flows from the winter storms. Incision and streamflow were probably generated from the high intensity summer storms, which generated more effective runoff.

With increasing aridification into the Holocene, climate and vegetation became similar betweenthe two areas and sediment production and delivery differences became less pronounced. Thelakes dried up and in most cases the reduced vegetation cover associated with the climatechange enhanced effective runoff generation, leading to erosion of slopes and fluvial trenchingof the upper parts of fan systems in both areas. This has led to the progradation of many of thefan systems, partly encouraged by associated drops in the lake (base) levels as the lakes dried up(Harvey et al. 1999).

Relevant reading

Harvey, A.M., Wigand, P.E. & Wells, S.G. (1999) Response of alluvial fan systems to the late Pleistocene toHolocene climatic transition: contrasts between the margins of pluvial Lakes Lahontan and Mojave, Nevadaand California, USA. Catena 36, 255–81.

Jannik, N.O., Phillips, F.M, Smith, G.I, et al. (1991) A 36Cl chronology of lacustrine sedimentation in thePleistocene Owens River system. Geological Society of America Bulletin 103, 1146–59.

ESC05 9/9/06 2:43 PM Page 178


latitudes suggests that the extent and location ofarid and semi-arid regions have changed (Fig.5.4). Generally greater aridity occurred in tropicallatitudes during glacial periods in high latitudes.In mid-latitudes altitude played an importantrole, with greater aridity in non-glaciated low-land areas such as the Mediterranean duringglacials, and relatively greater humidity duringassociated interglacial periods. Thus changes inrelative aridity in these regions would have ledto an expansion of areas susceptible to aeolianactivity. It has also been suggested that increasedwind speeds would have affected some areas dur-ing glacial periods, further exacerbating aeoliansediment movement. This in part is reflected inthe extent of relict ergs to be found in areas suchas Australia and the Sahara.

5.4.3 Short term time-scales: storminess

For shorter time-scales, such as the individualstorm event, it becomes apparent that the sim-pler models of (net) water and sediment move-ment on slopes considered for longer time-scalesare too simple. The problems are in monitoringthe more complicated detail of storm events.Much work has been undertaken using rainfallsimulation in the field, but there is much debateon what is a suitable scale for these plots to be of use. There is also much debate about the pre-cise controls of sediment and water movement,and it would appear that the control is scale-dependent. In order to consider overall sedimenttransfer this section will concentrate on the hill-slope scale (Case Study 5.3).

Case study 5.3 Runoff and sediment movement on a hillslope scale, results of the MEDALUS andIBERLIM European Union projects

Partly as a response to growing concerns over desertification in the drier (arid to semi-arid)regions of Europe, the European Union supported research into sediment and water transferacross the region, at the catchment and hillslope scale, in the 1990s. Examples of two of theseprojects are MEDALUS (Mediterranean Desertification and Land Use) and IBERLIM (ErosionLimitation in the Iberian Peninsula). The MEDALUS project was primarily involved in theexamination of uncultivated sites, whereas IBERLIM was focused on cultivated (afforested) sites.The field results and field experimentation at the plot scale will form the case study into shortterm (individual storm) significance in sediment transfer and runoff in managed (IBERLIM,central Spain) and non-managed (MEDALUS, south-eastern Spain) semi-arid catchments.

The MEDALUS project developed one of its erosion monitoring sites in the Rambla Hondain the early 1990s. The experimental catchment is a first-order catchment developed on mica-schists and consists of bare hillslope components in the upper parts and alluvial fan materialderived from the mica-schist in the lower part. In general the hillslope element was acting as asource of hydrological pathways (surface runoff) and sediment movement, and the alluvialmaterial at the base as a sink. At the individual storm-event scale, however, it was found thatwidespread transfers of water and sediment were unusual and of short duration, even in extremerainstorm events. In the latter events although local runoff generation may be high on slopelengths of 10 m, they decrease dramatically on longer, 50 m length hillslopes (Puigdefabregaset al. 1998). On the slope surface runoff occurred by infiltration excess in the early part of the individual events, but was dominated by local subsurface saturation of upper layers of the soil. Across the hillslope surface mosaics of plant clumps and bare earth became important.Runoff was generated from bare earth areas but lost to vegetation clumps, which acted as sinks.Runoff was thus laterally discontinuous, even in the larger observed areas in storm events. Onthe hillslope scale the only time connectivity between hillslope and channel elements can occuris when the saturation of the shallow surface layer occurs across the entire slope to enable the

ESC05 9/9/06 2:43 PM Page 179

180 ANNE MATHER

Headward development of'v' shaped axial gullyinitiated by overland flowon low permeabilitymaterials possiblyrelated to periods offormer agriculturalactivity.

Stage 1

Stage 3

Headward extension andgully incision expose piping-susceptible sediments tooverland flow generatedfurther up gully. Dessicationcracks provide water entrypoints with the developmentof vertical inlet pipes supplyingsubhorizontal collector pipes,which follow the principal gullyaxis. Bulbous form related topiping, pipe collapse andslumping of saturatedsediments.

Stage 2

Piping-susceptible horizons

Non-susceptible horizons

Matorral scrub

Vertical 'funnel' pipe

Pipe outlet

Overland flow

Subsurface flow alongsubhorizontal collector pipes

appr

ox. 7

0 m

approx. 100 m

approx. 50 m

Disintegration of pipe systemby pipe enlargement, collapse

and continued erosion leaves fluted susceptible horizons exposed on gully

sidewalls. Small vertical funnel pipes and remnants of dissected pipe network. Axial gully

development on sideslopes of main gully. Patchy vegetation colonization.

Case Fig. 5.3 Gully evolution and piping at the IBERLIM study site in central Spain. (From Ternan et al. 1998.)

ESC05 9/9/06 2:43 PM Page 180


connection of the hillslope elements to the first-order channel, and this occurs extremely rarely(Puigdefabregas et al. 1999), and was not witnessed during the project.

The IBERLIM managed site covers a 3.68 ha gully catchment in central Spain, located on theRaña Formation (Pliocene alluvial sediments). Prior to the 1940s the area was used as agriculturalland, which then became abandoned and taken over by Cistus mattoral vegetation (Ternan et al.1996a). In an attempt to combat erosion problems in the area an afforestation programme wasundertaken and in the 1970s the catchment was bench terraced and planted with Pinus. TheIBERLIM team began monitoring in 1992, on a variety of scales (Williams et al. 1995) from (i)gully (3.5 ha), (ii) erosion plot (10–21.5 m2) to (iii) rainfall simulation plot (1 m2). The researchdetermined that the bulk soil had an overall low aggregate stability (Ternan et al. 1996a), whichwas exacerbated locally by high clay content (especially if expandable clays were present). Notall sediment and water erosion occurred at the surface, with a good deal of erosion occurringthrough subsurface pipes, which developed in layers within the sediment where dispersion,shrinkage and swelling were prolific (Case Fig. 5.3). This again coincided with clay horizonslacking in coarse material (typically less than 40% sand). The sediment dispersion was found tolink closely with the chemistry of the porewater, soil chemistry and clay minerals. Horizons withsodium absorption ratios (SAR) of more than 0.4 and exchangeable sodium percentage (ESP)of more than 1.5 were found to be susceptible (Ternan et al. 1998). These horizons had beenexposed with the excavation of the bench terraces within the woodland (Ternan et al. 1996b).These areas were thus sources of runoff and associated surface erosion early in a rainstormevent. Continuous runoff pathways were more prolific under wet weather conditions. It wasdemonstrated that to manage erosion in such areas, and also to assist in the minimization offlooding, both the overall wetness threshold at which runoff is generated throughout the catch-ment should be considered together with the spatial mosaic of runoff-generating source areasand sinks (Fitzjohn et al. 2002). If sinks could be used to disrupt continuous hydrological flowpathways both slope sediment and water transfer could be minimized.

The MEDALUS and IBERLIM studies serve to emphasize the mosaic nature of runoff generationon hillslopes in arid environments. The resulting data demonstrate how human interventioncan exacerbate problems by increasing linkages of source areas on slopes and indicates that aknowledge of the linkages could be used to mitigate runoff problems by using sink areas to breakup flow pathways.

Relevant reading

Fitzjohn, C., Ternan, J.L.,Williams, A.G., et al. (2002) Dealing with soil variability: some insights from landdegradation research in central Spain. Land Degradation and Development 13, 141–50.

Fitzjohn, C., Ternan, J.L. & Williams, A.G. (1998) Soil moisture variability in a semi-arid gully catchment:implications for runoff and erosion control. Catena 32, 55–70.

Puigdefabregas, J., Sole, A., Gutierrez, L., et al. (1999) Scales and processes of water and sediment redistributionin drylands: results from the Rambla Honda field site in Southeast Spain. Earth Science Reviews 48, 39–70.

Puigdefabregas, J., del Barrio, G., Boer, M.M., et al. (1998) Differential responses of hillslope and channel elements to rainfall events in a semi-arid area. Geomorphology 23, 337–51.

Ternan, J.L., Elmes, A., Fitzjohn, C., et al. (1998) Piping susceptibility and the role of hydro-geomorphic controlsin pipe development in alluvial sediments, Central Spain. Zeitschrift Für Geomorphologie 42, 75–87.

Ternan, J.L., Williams, A.G., Elmes, A., et al. (1996a) Aggregate stability of soils in central Spain and the role ofland management. Earth Surface Processes and Landforms 21, 181–93.

Ternan, J.L., Williams, A.G., Elmes, A., et al. (1996b) The effectiveness of bench-terracing and afforestation forerosion control on Raña sediments in central Spain. Land Degradation & Development 7, 337–51.

Williams, A.G., Ternan, J.L., Elmes, A., et al. (1995) A field study of the influence of land management and soilproperties on runoff and soil loss in central Spain. Environmental Monitoring and Assessment 37, 333–45.

ESC05 9/9/06 2:43 PM Page 181

182 ANNE MATHER

Recent research has highlighted the signific-ance of heterogeneities in slopes, which generatea ‘mosaic’ of hydrological characteristics. Thesemosaics can be considered as ‘pixels’ in a digitalpicture, where each pixel has a different indi-vidual response to a rainstorm event, even if the intensity and duration of the event were con-stant across the entire area (which is unlikely).The hydrological response of each pixel will bedetermined by its ability to store water. This willbe determined through a range of factors such asporosity, permeability, organic matter content,slope angle, etc. As a function of these charac-teristics some pixels will act as ‘sources’ andsome will act as ‘sinks’ (Fig. 5.18). The sourceswill respond to the rainstorm event and gener-ate runoff. This runoff will continue downslope

until it hits a sink. Runoff will be continuousonly where the source cells dominate the sinks(Fig. 5.18). As rainfall inputs vary then so willthe number of cells acting as sources or sinks.The amount of sediment carried by this run-off will be dependent on the ‘erodibility’ of thesediment and the ‘erosivity’ of the rainfall. Thelatter will be determined mainly by the rain-storm characteristics (rainfall intensity and dropsize), but may be modified by the intervention of vegetation. The erodibility of the sedimentswill be determined by the aggregate stability,which in turn is affected by the soil structure(amount of binding organic matter or presenceof dispersing agents such as sodium salts). Thus,the nature of the sediment characteristics acrossand into the slope, combined with the slopemorphology and cover characteristics, added to the spatial variability of an individual stormevent will generate a spatially varied movementof sediment and water across a slope. This canbe further complicated by the intervention ofhumans, modifying the local hydrological characteristics (Case Study 5.3). All factors considered it is not difficult to see why the geo-logical concept of a ‘sheetflood’ rarely, if ever,occurs in the natural environment!

5.5 ANTHROPOGENIC IMPACTS

Humans tend to affect the landscape on thesame time-scales as climate. Thus, the impact of humans is often difficult to unravel from thenatural controls discussed above. This sectionwill illustrate the significance that these impactsexert on controlling the modern landscape,which in turn has implications for how environ-ments can be managed sustainably. The world’sarid environments have long been occupied by humans. In more recent times, however,improvements in technologies of water manage-ment, improved healthcare and hygiene havemeant an increase in population of these regions.In many cases this may be exacerbated by the availability of mineral deposits. The sectionexamines the impact of humans on the erosion,transport and deposition of sediment in deserts.

A B C D E F G HA B C D E F G H

High

Zero

Flow in channel at slope base

(a)

(b)

A B C D E F G H

Slope top

Slope base

A B C D E F G H

Fig. 5.18 Conceptual model of the mosaic of surface runoffpathways on a hypothetical slope (a). Each cell represents asource (dark) or sink (light). If it is assumed that (i) rainfall iseven across the slope and (ii) that source and sink capabilities percell (pixel) are equal, the contribution of runoff from the slopemosaic to the channel is displayed in (b).

ESC05 9/9/06 2:43 PM Page 182


5.5.1 Water resources and salinization

Salinization is the process of accumulation ofsoluble salts in the upper parts of soil horizons.Secondary salinization (where salinization isenhanced by human intervention) of soils in aridareas is a major problem. This is particularly so where groundwater levels are close to the sur-face and salts can be drawn into the upper partsof the soil where they naturally accumulate. Asimple removal of vegetation cover may be allthat is required to modify the hydrology of thesoil (raise the water table) and place it in dangerof salinization in an arid area. For example, inAustralia native evergreen forest removed bysettlers increased the salinity of over 200,000 hato the point where they were no longer produc-tive (Mackay 1990). Secondary salinization morecommonly occurs through excessive irrigation.This is not surprising given the rapid increase inirrigation globally since 1800. In some countries(e.g. Egypt) nearly 100% of land in agriculturaluse is irrigated (Rhoades 1990). Approximatelyhalf of all irrigated areas suffer from problemsof secondary salinization (Rhoades 1990).

Salinization can lead to cracking and puffingof salt-affected soils (Mabbutt 1986), affectingthe hydrological characteristics of the soil. Wherea salt crust or accumulation forms runoff can be exacerbated. The salts will also exacerbateweathering rates, producing abundant fine mater-ial which may then be deflated by wind action,leading to the formation of ‘salt scalds’ (Mabutt1986). In hyper-arid areas where salts natur-ally occur in the soils, for example in the form of gypcretes (Table 5.5), irrigation may lead to rapid dissolution and collapse of the soil surface, affecting the urban area. As salt is a dominant weathering agent in arid landscapes(Doornkamp & Ibrahim 1990; section 5.2.1.3)it can also lead to significant impacts on engineer-ing structures and urban areas (section 5.6.3).

5.5.2 Agricultural resources and soil erosion

Soil erosion by water has been recognized re-cently as a major contributor to erosion in aridlands. Natural vegetation clearance and increased

populations have accelerated rates of soils erosion.Regions bordering the Mediterranean Sea havesome of the highest erosion rates in the world(Woodward 1995), with some 1000 t km−2 yr−1 ofsuspended sediment generated. Of this, 70% canbe attributed to anthropogenic agents (Dedkov& Mozzherin 1992). Within this region somecountries have implemented the policy of reduc-ing the number of people involved in agricultureand replacing them with mechanization. Suchcountries have been experiencing acceleratedrates of erosion (Morgan 1994). In some casesthis is related to the abandonment of agriculturalterraces (e.g. Millington 1989). Mismanagementof the landscape can also lead to alterations inthe hillslope hydrology, which can have an impacton runoff pathways and exacerbate soil erosionrelated to surface runoff (Case Study 5.3, theIBERLIM example).

Soil erosion by wind is another common phenomenon. It is particularly common in areaswhere water resources have been used for irrigation, leading to a lowering of the affectedwater bodies and exposure of lake sediments todeflation. Wind erosion is further exacerbatedwhere secondary salinization inhibits vegetationgrowth, thus limiting the erosional protectionbenefits of vegetation cover. In other cases, soilerosion is related to overgrazing of land or over-production in years of drought, which can leadto a loss of arable plant protection. Deflationcan become a major issue where this coincideswith fine-grained sediments such as glacial andriver silts or loess.

5.5.3 Mineral resource extraction

Arid environments are often locations of mineralresources such as the diamonds of Namibia orthe salts and nitrates of the Atacama of northernChile. The impact of this mining on the environ-ment is highly variable. In the hyper-arid climateof the Atacama of Chile extensive nitrate min-ing occurred across the Central Depression. Themining at its peak provided almost 50% of theincome for the Chilean Government (Wisniak &Garces 2001). Most of the mines were abandonedin the 1930s and now litter the Central Depression

ESC05 9/9/06 2:43 PM Page 183

184 ANNE MATHER

in a fossilized landscape, but their impact uponsediment systems is considered to be minimal.

In contrast the landscape response to aband-oned mining is more dynamic in semi-arid envir-onments. The Rio Tinto fluvial system in theProvince of Huelva, south-west Spain has a meanannual rainfall at Huelva of around 500 mm yr−1

(Instituto Nacional de Meterologia). The annualdischarge ranges from close to zero in the summerto 11 m3 s−1 in the winter. The lowest 20 km of the catchment is tidal, with a mean tidalrange of 2.2 m. The waters of the river are acidic(typically pH 2.0–2.5), but become less acidicwith the addition of Atlantic waters on flood tides.The Rio Tinto lies within the Iberian Pyrite Belt,one of the largest deposits of its type in the world.The main rich ore body is 5 km long, 750 m wideand 40 m deep. It contains abundant Fe, Cu, Zn,Pb, Ag and Au. Mining of the sulphide depositshas been occurring with increasing efficiency for some 5000 years, ceasing in the late 1990s.The impact of improved mining technologies andefficiencies can be demonstrated by comparingthe amount mined in the Rio Tinto catchment inmillion metric tons (values taken from Davis et al.2000). The Tartessans removed 3 million tonnesover 4200 years (5000–2800 yr BP); the Romansremoved 24.5 in 200 years (2000–1800 yr BP);and the British (Rio Tinto Zinc) re-moved 1600in the last 200 years. These mining operationshave liberated abundant suspended sedimentand solutes into the river system. Cores of 1 mdepth taken throughout the catchment from thechannel margins and dated with 210Pb and 14Cshow overall rates of sediment accumulationwithin the catchment of 0.3 cm yr−1. High levelsof pollutants (e.g. Cu at 970 ppm for 840 yr BPand 2466 ppm at 3640 yr BP) were found to pre-date the RTZ opencast mining (Davis et al. 2000).

5.6 LIVING WITH DESERTS: AN ENGINEERING PERSPECTIVE

Deserts provide engineering challenges where thearid environment may have to be traversed orinhabited. This section will consider the engineer-ing management and remediation of sediment-based hazards in desert environments today.

5.6.1 Aeolian hazards

The physical detachment, entrainment and trans-port of solid particles has been enhanced in manyregions where the erodibility of the soils has beenexceeded by the erosive agents, which includehumans. Soil erosion exacerbated by humanshas been recorded from the time of Plato (fourthcentury bc) to more recent history (e.g. O’Haraet al. 1993). Aeolian removal of sediments is a major problem in arid environments. Winderosion affects 39% of arid areas susceptible tohuman degradation (UNEP 1992). Agriculturalmismanagement can lead to severe erosion andsources for available dust can also be exacerbatedby human activity, especially where major waterdiversions take place. Mitigation of these pro-cesses ranges from physically removing or limitingthe source area of the sands, building protectivebarriers and building to reduce the impact ofstructures (Table 5.6).

5.6.1.1 Sand dune encroachment

Arid environments are characterized by sparseor no vegetation. Thus free, mobile sand dunes(section 5.3.4) are common in these areas. Dunesoils have proved popular for agriculture insome regions, as they are less prone to salin-ization than other areas. These resources can beeasily overexploited so that vegetation is lost.Once the vegetation is removed, moisture loss isincreased and restabilization unlikely. In someareas this is further exacerbated by quarry-ing, for example in Kuwait, for urban buildingprojects. Thus, partially stabilized dunes maybecome reactivated. The drifting sand becomesa problem for road transport and can bury settlements.

Urban areas themselves pose a problem in aridenvironments as they offer an obstacle to sandmovement. In addition, associated urban activ-ities may exacerbate the magnitude of aeolianproblems through the loss of vegetation, etc.Sand and dust problems are thus prolific in theliterature on desertification. Areas dominated bysand dune encroachment experience the burialof structures and transport links, and features

ESC05 9/9/06 2:43 PM Page 184


such as pipelines, which are purposely exposedfor maintenance, can become inaccessible.

5.6.1.2 Soil erosion

Loss of topsoil through deflation will inhibit fertility, and can undermine structures throughscouring of footings. This can affect anything fromtelegraph poles to pipelines, railway sleepers androads (Cooke et al. 1982). Typically once thesand source has been depleted, and an armouredsurface has been reached, e.g. a gravel lag, orwhere desert pavement predominates, then defla-tion ceases to become a problem. It will onlyremain a problem in areas where the sand supplyis continually renewed, e.g. in mobile dune fieldsor along unpaved roads. The Arizona Depart-ment of Transport (1975) reported an annualloss of silt and clay by deflation at 5–50 kg pervehicle per mile of unpaved track.

5.6.1.3 Sand abrasion

Transport of sand by aeolian processes leads to the problems of abrasion. Typically, abrasionis exacerbated over indurated surfaces such astarmac and concrete (Cooke et al. 1982). Pitting

(surfaces more than 55° to the prevailing winddirection) and fluting (surfaces less than 55°to the prevailing wind direction) can occur onbuildings, and telegraph poles may suffer sandblasting damage near their base. Glass becomesfrosted and paintwork may be damaged on both buildings and vehicles. Air filters are alsoaffected. In addition, visibility can be impairedby sand storms. Dust storms have been reportedto create many problems including the trans-portation of pathogens, suffocation of cattleand interference with reception on transmittingand receiving devices.

5.6.2 Water hazards

Despite the arid nature of drylands, water is one of the main hazards in these regions. From an engineering perspective ephemeral aridstreams provide a challenge. Peak sediment andwater yields are commonly associated with thesemi-arid environment, where drainage densities are best developed (wetter areas have channeldevelopment inhibited by plants and drier areashave insufficient frequency of events to main-tain channels). Some authors (e.g. Farquarson et al. 1992) suggest strong similarities in flood

Table 5.6 Mitigation of aeolian hazards. (Summarized from Cooke et al. 1982.)

Method

Avoidance of problem areasRemoval of dunesVegetational Natural recoverystabilization Artificial recovery

Surface Waterstabilization

Lag

OilChemical spraysWood cellulose fibres

Fences WindbreaksDiversion fences

Impounding fences

Architectural planning

Comments

Best approachExpensive, but successful for small dunes. Need to be maintained.Protects the main areaSuccessful but expensive, and needs to be maintained. Careful plantselection is required to survive the hydrological conditionsNeeds to be maintained. In some instances mineral precipitationassociated with the evaporation of the water may enhance surface stabilityMaterial greater than 2mm diameter used as an artificial pavement. Canbe costly to transportSuccessful but unsightlyThe chemical crust generated may be fragile and expensiveSprayed as slurry with water fertilizer, grass seed and asphalt or emulsion.Fragile but cheapCan be living or artificial. Can be expensiveSet at an acute direction to the wind. Expensive and require maintenance(removal of lee sand)A series of fence increments, which extend the life of the barrier fromburial. Successful and relatively cheapMinimizing the impact of housing on wind flow and sand accumulation

ESC05 9/9/06 2:43 PM Page 185

186 ANNE MATHER

frequency between ephemeral rivers in arid zones.Other data (e.g. McMahon 1979), however,suggest that specific peak annual dischargesvary considerably.

A key problem in many arid areas is the slopeerosion of natural and artificial cuts principallythough gullying. This can lead to rapid erosionof cut surfaces. Mitigation of such factors typic-ally takes one of two approaches:1 reduction of the erosivity of the rainfall byreducing gradients of slopes, and where possiblekeeping slope lengths and hence catchment areasas small as possible;

2 reducing the erodibility of the sediments byprotecting the surface with vegetation or arti-ficial stabilizers.

In the main flood zone associated with largergully systems, alluvial fan systems and ephemeralriver channels, flash flood damage is the majorhazard. Although canalization can be used to con-strain flood events, the most successful approachis to avoid high risk areas such as ephemeralchannels on alluvial fans and river systems. Aridrivers are associated with large amounts of scourand fill in individual events (Fig. 5.19). Theseprocesses have significant implications for the

(a)

(b)

Fig. 5.19 Deposition and scourimpacts of individual flash flood events.(a) Coarse sediment deposition(railway line has been excavated fromdeposits at base of photograph) fromthe 1991 Antofagasta flood in NorthernChile. (Photograph courtesy ofGuillermo Chong, University ofAntofagasta.) (b) Bridge removal byscour in a bedrock tributary of the Draariver, southern Morocco.

ESC05 9/9/06 2:43 PM Page 186


loss of housing and transport routes throughburial and erosion.

Ultimately, in many river systems, the sus-pended sediment will end up in a playa or lakesystem. Reservoirs in arid areas are thus subjectto high rates of sedimentation, so that their half-lives are considerably reduced. For example, theTarbella Reservoir on the Indus lost 6% of itscapacity within 5 years of construction (Ackers& Thompson 1987). Thus, siltation is a primaryconcern for reservoirs in arid zones, with miti-gation taking the form of slope stabilization byvegetation, check dams to retain sediments ingully systems and maintenance of the reservoirthough flushing.

Another potential problem associated withwater in arid environments is the developmentof karst and pseudokarst. The lack of watermeans that deposits with a low unconfined com-pressive strength such as gypsum can becomescarp-forming caprocks. Arid environments alsocommonly supply the necessary geological andsoil-chemistry conditions (e.g. unconsolidatedsediments, abundant dispersive salts such assodium) for the subsurface flow of water, lead-ing to the development of subsurface piping or‘pseudokarst’. Where these features develop theyoffer problems for construction and, as they are liable to change though collapse and naturalexpansion (Fig. 5.20), can lead to the destabil-ization of existing structures.

5.6.3 Salt hazards

Salt weathering is a major feature of deserts(Doornkamp & Ibrahim 1990). Salt weatheringaccelerates the breakdown of source rocks tosilt-grade material (Goudie 1984) thus makingthem more susceptible to other sediment erosionagents such as wind and water, and exacerbat-ing dust problems. It also attacks the fabric ofhouses, roads (Cooke et al. 1982; Doornkamp& Ibrahim 1990) and important archaeologicalfeatures such as the Sphinx in Egypt. In manycases accelerated salt weathering may be directlyrelated to increased salt in the local environmentas a function of irrigation (Goudie 1977) orbuilding-site leakages. The main impacts of salt

weathering in urban areas are through (i) chem-ical alteration (e.g. corroding reinforcing bars inconcrete and chemically altering concrete) and(ii) volumetric changes due to salt weathering(e.g. leaching of salts from foundations causingsubsidence, or hydration leading to expansionand ground heave).

5.7 FUTURE ISSUES

The population of Earth has increased from some2.5 billion people in the 1950s to 6.1 billion inthe year 2000. This growth in population createsgreater pressure to use resources in arid environ-ments and increases the risk of desertification.In all it has been calculated that there are some49 million square kilometres on Earth at somerisk of desertification (Eswaran et al. 2001).Most of this degradation is associated with un-sustainable land management practices, whichhave an impact on the soil and vegetation and

Fig. 5.20 Megapipe developed in gypsiferous Quaternarydeposits, south-east Spain.

ESC05 9/9/06 2:43 PM Page 187

188 ANNE MATHER

are exacerbated during periods of drought, suchas in the Sahel in the 1970s. However, there are problems in distinguishing natural drylandecosystem variability from longer term naturalfluctuations in climate. For example, Lamprey(1975) analysed data from 1958 to 1975 andconcluded that as a function of overgrazing theSahara was marching northwards at 5.5 km yr−1.However, new data sources and approaches (e.g.Hellden 1988; Le Houérou et al. 1988; Tuckeret al. 1991; Hulme 1992) indicate that the de-gradation (desertification) observed by Lampreywas in fact attributable to natural changes inweather conditions (i.e. drought), comparing wetyears (1958) with dry years (1975). Quaternaryrecords over the past 40 kyr show that the Saharahas naturally ranged in climate from humid/subhumid to hyper-arid/arid (Fig. 5.4). Over the Holocene Epoch fluctuations in aridity (4.5–3.5 ka) have been attributed to the socio-economiccollapse of several cultures (Petit-Maire & Guo1998). It is now accepted that dryland vegeta-tion systems are dynamic and far from fragile,showing adaptation and resilience to environ-mental change. So although advancing sand dunesdestroying agricultural land may be true locally,the issue of land degradation may be overestim-ated and confused with natural environmentalvariability (Thomas & Middleton 1994). Theseconcepts are particularly pertinent in the con-text of future global climate change, which mayalter the spatial pattern and extent of drylandsand the severity of their associated problems forhuman habitation. Thus there is a need to betterunderstand the dynamism of drylands and howthey respond to environmental change.

To assess the extent to which current trends indryland development are attributable to longerterm controls such as climate, there is a needboth to exploit the data stored in ancient recordsand to develop higher resolution Holocene records(Fig. 5.4). These provide valuable data stores on the past extent of arid realms. For example,aeolian dune sediments provide information onpast wind regimes. The identification of aeoliandune sediments in palaeochannels may help usidentify the presence of periodically drier condi-tions in the ancient record. Palaeolakes offer a

wealth of information on former high and lowstands reflecting changes in the hydrology of the lake basin. Rivers and alluvial fans may yieldinformation on changes in climate relating to thegeneration of effective runoff (Case Study 5.2).Caves, karst, tufas and groundwater provideinformation on wetter versus drier periods withinthe climate record. Floral and faunal changesmay indicate landscape and climate change, suchas the pack-rat middens of the south-westernUSA (Case Study 5.2). Archaeological evidenceof human behaviour may reflect changes in localand regional climate.

Where absolute dates can be applied to theabove data then rates of change can be determinedwithin the arid realm. The advent of new methodsof dating enables higher resolution chronologiesto be established, allowing us to better correlateand understand the larger scale and longer termimpacts of aridity. The responses to climatechange in arid lands can be complicated and, asdemonstrated in this chapter, wetter climates donot necessarily mean more erosion. In contrastdrier, more seasonal climates may tend to lead togreater erosion as a direct function of their effecton vegetation cover and, through that, erosion.It is also clear that an understanding of the long-term evolution of landscapes (Case Study 5.1)may improve understanding of both the spatialand temporal process dynamics within an area,and that if we are going to manage these land-scapes to minimize the impact of human inter-vention these need consideration.

Prediction of future events relies on an under-standing of both the palaeoenvironmental recordand the contemporary record. As demonstratedin the case studies in this chapter, however, isolating cause is far from straightforward, evenin the contemporary. In addition there are theadded difficulties of the differing time-scales on which processes operate. For example theform of dryland valleys, which formerly hasbeen explained by changes in surface erosionaland depositional factors, may be more stronglyrelated to other factors such as groundwatersapping and deep weathering (Nash et al. 1994)in some arid settings, which operate on differenttemporal scales.

ESC05 9/9/06 2:43 PM Page 188


Field experimentation on a range of scales has been attempted to try and deal with some of the above spatial and temporal issues. In the1990s the EU-funded experimental field studiesIBERLIM and MEDALUS (Case Study 5.3) generated field data that were used to generatecomputer simulation models (Kirkby et al. 1998)of surface runoff, vegetation and erosion andtheir interplay through space and time on both ahillslope (MEDALUS model) and large, up to5000 km2, catchment scale (MEDRUSH model).These were calibrated with the collected fielddata and are being used to explore future trendsunder different scenarios. All modelling, how-ever, is fraught with difficulties of scale. Globalclimate models (GCMs) use cells of < 0.5° latitudeand longitude, and yet previous work has shownthat detailed knowledge is needed of a widerange of variables, much smaller than this cellsize (such as rocky surface conditions and cover,Yair 1994), to have any hope of accurately pre-dicting the magnitude and size of change.

Debates rage on prediction of future climatechange and whether global warming or cool-ing will prevail. Predictions from the generatedGCMs vary widely. Integration of the data from three of the main GCMs for doubledatmospheric CO2 levels indicates that southernEuropean drylands (bordering the Mediterranean)will have increases in summer temperature of + 4 to + 6°C in the summer, and + 2 to + 6°C inthe winter, with associated overall significantdecreases in precipitation and overall avail-able soil moisture (Williams & Balling 1995).The Hadley Centre for Climate Prediction andResearch (UK) provides a summary of some of thevarious types of models available through linkson www.metoffice.com. These predictions (to

the year 2100) point towards raised temperatures,lower precipitation and lower soil moisture ingeneral, for the zone 30°N–30°S of the Equator.This zone contains many of the world’s aridenvironments (Fig. 5.2). In areas such as theAtacama Desert of South America impacts ofglobal climate change on the duration, magni-tude and frequency of El Niño and La Niña needto be understood, as these phenomena are signi-ficant for rainfall generation in such regions.

Dryland landscapes and their associated vegetation systems, which help control rates oflandscape change, are naturally dynamic andhave a degree of resilience to environmentalchange. These systems thus have some naturalresistance to human intrusion. With a burgeon-ing global population, however, our interventionwith the arid environment will only increase asdemand for resources increases, particularly if, as predicted by GCMs, the spatial extent of drylands expands/intensifies in response tofactors such as global warming. Thus there is a need to address the issues of sustainability insuch environments in order to avoid currentlylocalized impacts becoming more global. Thismay require embracing remote sensing tech-nology and integrating it with field studies to monitor dryland environments and the natureof changes within them. These data, togetherwith longer term Quaternary data, can be usedto try and better understand the complexities of dryland environments and their response toenvironmental change. If, however, this know-ledge is to be successfully implemented in thesustainable management of dryland systems we need to ensure that scientists communicatewith the decision makers more effectively(Thomas 1997c).

ESC05 9/9/06 2:43 PM Page 189

6.1 INTRODUCTION AND DEFINITIONS

6.1.1 Introduction

The urban environment is one that is of increas-ing importance globally, with implications forboth hydrological and sedimentological systems.It has been estimated that 50% of the Earth’spopulation live or work in an urban environ-ment (United Nations 2003) and this percent-age is predicted to increase. Urbanization dates back several thousand years, with developmenttaking place primarily along major rivers actingas waterways. It was not until the IndustrialRevolution in Europe, and later throughout otherparts of the world, however, that the growth of significant urban areas took place. The verynature of urban land surfaces and waterwaysresults in unique hydrological, sedimentologicaland atmospheric environments, leading to a widerange of specific management and sustainabilityissues, many of which are not encountered innatural environments. The application of processmodels derived from observation and measure-ment of natural systems to urban environmentsis, therefore, of limited value. The issues of environmental pollution and sustainability haveresulted in a plethora of research into the qualityof urban air, groundwater and, to a lesser extent,surface water. Urban sediments, and the role of the urban environment as a sedimentary sys-tem, have been largely neglected. It is increasinglybeing recognized, however, that particulates inurban environments are a major factor in humanhealth, through their impact upon air quality,waterways and biodiversity, through the role ofsediments as vectors for contaminant transfer and

6Urban environments

Kevin Taylor

reservoirs for contaminant storage. As a result,our knowledge base on urban sedimentology, andthe interactions with hydrology, air quality andbiodiversity, is rapidly increasing. This chapterdescribes urban sedimentary environments, thesources of sediments to these environments, the physical and chemical characteristics of theresulting accumulated sediment, and discussesthe impacts of intrinsic and extrinsic processesupon these sediments. Finally, the sustainablemanagement options for urban sediments andtheir impacts will be discussed.

6.1.2 Definitions

The term urban is used widely throughout bothsocial and scientific literature, and has come to mean different things to different people. A definition suggested here for urban areas is‘those areas where the ecosystem is significantlymodified by human settlement and associatedactivities; they are characterized by a uniquemodification of the physical, chemical and bio-logical environments, resulting from the con-struction of buildings on a large scale’. One key element of the urban environment that dis-tinguishes it from the environments discussed in the other chapters of this book is its highly engineered nature. Such engineering results inland surfaces with distinct physical and chem-ical properties. Although natural environmentshave inevitably been altered by anthropogenicactivity, these have not resulted in the signific-ant, and wholesale, change in fundamental pro-cesses of sediment and water movement andaccumulation seen in urban environments. Theseinclude increased storm peaks and shortened

ESC06 9/9/06 3:01 PM Page 190

URBAN ENVIRONMENTS 191

lag times, with greater opportunity for scouring,erosion and transport of sediment. Furthermore,the extensive presence of transport and, to a lesserdegree, industry in urban environments leads to increased contaminant sources and enhancedpollution pressures.

The term urban sediment is commonly usedand is used in a range of different contextsthroughout the literature. In the past urban sediments has been a term commonly used torefer to sediment accumulation on street surfaces,although in this chapter the term is used in abroader manner to mean any sediments presentwithin the urban environment. They, therefore,comprise sediments accumulating on street sur-faces, in gully pots and sewer systems, and inreceiving water bodies (rivers, canals, docks andlakes; Figs 6.1 & 6.2).

The term road dust has also been used exten-sively in the past, but the term dust implies fine,respirable material (i.e. grains less than 10 μm).Numerous studies have shown that street sedi-ment is composed of a full range of particlesizes, commonly biased towards coarse materialand, therefore, the use of the term dust is notappropriate. The terms street sediment or roadsediment have also been used in the past, butwith no formal definition. More recently theterm road-deposited sediment has been used(e.g. Sutherland 2003). This term is favouredhere, owing to its clear descriptive nature, and

will be adopted throughout this chapter. Road-deposited sediment (RDS) can be thought of as predominantly subaerial in nature and con-trasts with the subaqueous (or aquatic) sediments that are present within urban rivers, canals anddocks. Sediments accumulated within gully potsand sewer systems are neither wholly subaerialnor subaqueous, but experience both conditionsepisodically dependent on weather conditions.These are, therefore, described separately in thischapter from other sediment types.

6.1.3 Historical development of urbansedimentology

The study of urban sediments from the perspec-tive of environmental sedimentology is a youngone. Research into urban particulates originatedout of concern for pollution and human health,and focused on road-deposited sediment (RDS).Early work focused on the levels of lead in RDS,as this is a major vector for lead consumption by children through hand-to-mouth activities.This research documented the high levels of lead in RDS, compared with crustal abundance(Farmer & Lyon 1977; Duggan & Williams1977; Harrison 1979). The importance of RDSas a vector for Pb uptake was shown by the cor-relation between RDS-Pb and blood-Pb levels inchildren (Thornton et al. 1994). This was streng-thened by epidemiological studies that showed a

Lakes

Docks and canals

Rivers

Streetsurfaces

Sewers

Gully pot

Fig. 6.1 Schematic diagram of the urbansedimentary environments covered in this chapter.

ESC06 9/9/06 3:01 PM Page 191

192 KEVIN TAYLOR

clear link between high lead intake and intel-lectual development in children (Needleman et al. 1979). Partially as a direct result of thispioneering research, the use of leaded petrol was phased out in Europe. Following on fromthis, the levels for a wide range of metal andorganics contaminant have been studied in RDS from around the world and this aspect ofurban sediment study is now well-established(section 6.3.1).

The pressures of water quality regulations,biodiversity issues and regeneration of water-side areas led to the second main area of urbansediment research – aquatic urban sediments. Ithas long been recognized that sediments act asthe major vector for the transport of contam-inants in most aquatic systems (see Chapter 1)as contaminants are preferentially associated

with the particulate phase. It has also been well-established that sediments act as a majorstore of contaminants (e.g. floodplains and saltmarshes). Such storage, and potential remobil-ization, is a major issue in urban rivers, canalsand docks, as a result of a legacy of unregulateddomestic and industrial discharge into thesewater bodies (Taylor et al. 2003). The study ofaquatic urban sediments has, therefore, becomean increasingly important area, not just as aresearch priority, but also for urban pollutionmanagement, sustainable river basin manage-ment and remediation. These studies have focusedon sediment and contaminant fluxes in urbanrivers, sediment quality in canals and docks, andunderstanding the short- to long-term pathwaysof sediment-bound contaminants through theurban environment.

RDS

(a) (b)

(c) (d)

Fig. 6.2 Photographs of the typical environments, and their characteristics, associated with urban sediments. (a) Street surfaces. (b)Accumulation of road-deposited sediment (RDS) around a traffic island on a street surface. Note the highly variable grain-size of thismaterial. (c) An urban river, showing the classic engineered and steep-sided nature of such rivers. (d) An urban dock, displaying surfacedebris accumulation due to prevailing wind directions.

ESC06 9/9/06 3:01 PM Page 192


6.2 SEDIMENT SOURCES AND SEDIMENT TRANSPORT

PROCESSES

6.2.1 Sediment and contaminant sources

Sediments within urban environments originatefrom a wide range of sources, both natural andanthropogenic. In this chapter sediment sourcesto subaerial environments (road-deposited sedi-ments) and subaqueous environments (rivers,canals/docks and lakes) will be considered separately. Although both environments receivea broadly similar range of source material, the differences in the physical and chemicalcharacteristics between the environments leadto contrasting intrinsic and extrinsic sourcesdominating.

6.2.1.1 Road-deposited sediment

Compared with sediment in natural environ-ments, road-deposited sediment (RDS) has a wide,and diverse, range of sources (Fig. 6.3). Sourcesare either intrinsic to the road surface, which are predominantly anthropogenic in nature, orextrinsic, which are predominantly naturallyderived. Intrinsic sources include vehicle exhaustemissions, vehicle tyre and body wear, brake-lining material, building and construction mater-ial, road salt, road paint and pedestrian debris.Extrinsic sources are soil material, plant andleaf litter, and atmospheric deposition.

There are very few studies that have attemptedto quantify the importance of these sources toRDS, partly as a consequence of the wide range

of sources, and partly due to the heterogeneousnature of most RDS deposits. Volumetrically, themost important component of RDS comes fromsoils and building material. Soil material, whichmay be derived from a range of distances, con-tributes both minerogenic and organic material.For example, a study by Hopke et al. (1980) recognized that soil material formed approxim-ately 75% of RDS. Building material contributesquartz sand, concrete and cement to RDS, andthese are believed to be relatively inert. In urbanareas undergoing extensive development, how-ever, the volume of building material input toRDS can be large, and may have important con-sequences for air quality, sediment volumes andparticulate-associated contaminants dischargedinto drainage systems and rivers.

Iron oxides are abundant in RDS and as aresult mineral magnetic analysis has been appliedto study source apportionment in these sediments.Mineral magnetic analysis can discriminate soil-derived iron oxide particulates from thosederived from fossil-fuel combustion (Oldfield et al. 1985), and as such has great potential inurban sediment sourcing. Application of mineralmagnetic analysis has shown the importance of non-soil derived material in RDS, includingvehicle and industrial combustion, constructionmaterial and asphalt abrasion (e.g. Beckwith et al.1986; Xie et al. 1999; Lecoanet et al. 2003). In astudy of RDS in Manchester, UK, Robertson et al.(2003) concluded that vehicular sources were theprimary contributor to iron oxide material (seeCase Study 6.1). Although the application ofmineral magnetic analysis has helped to identify

Constructionmaterial

Pedestriandebris

Atmospheric input

Building wear

Vegetation

Soil materialTyre wear Vehicle wear

Exhaustemissions

Road wearFig. 6.3 Schematic diagram of thesources of sediment comprising road-deposited sediments.

ESC06 9/9/06 3:01 PM Page 193

194 KEVIN TAYLOR

Case study 6.1 Road-deposited sediment, Manchester, UK

Road-deposited sediments (RDS) are a major component of the urban sediment cascade (see section 6.2.2) and are sourced from a wide range of material. An extensive data set on thephysical, chemical and mineralogical composition of RDS exists for the city of Manchester,north-west England (see Robertson et al. 2003). The sediments contain high metal concentra-tions, in most cases significantly above average crustal abundances. Lead ranges from 120 to645 μg g−1, Cu from 39 to 283 μg g−1 and Zn from 172 to 2183 μg g−1. Temporal analysis hasshown that Pb levels have decreased since the 1990s in response to the reduction in use of

0100200300400500600700

(a) (b)

Inner city Outer cityLocation

Pb

co

nce

ntr

atio

n (

µg/g

)

0%

20%

40%

60%

80%

100%

(c)

Sample

Oxidizable

Reducible

Exchangeable

SampleSample1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

0%

20%

40%

60%

80%

100%

0%

20%

40%

60%

80%

100%

Zn

Pb Cu

Case Fig. 6.1 (a) Box plot showing Pb levels in road-deposited sediment(RDS) in inner city and outer cityManchester. Note the higher levels of Pb in the city centre related to highervehicular activity. (Derived from datapublished in Robertson et al. 2003.) (b) Scanning electron microscopyimages of iron rich glasses, probablyproduced via hydrocarbon combustion.Note that these glass grains are sphericaland are approximately 50 μm in size.(Images courtesy of Davina Robertson).(c) Sequential extraction analysis ofmetals in RDS in Manchester. Zincshows a significant component in theexchangeable fraction, suggestingincreased mobility. (From Robertson et al. 2003.)

ESC06 9/9/06 3:01 PM Page 194


sources of iron oxides in RDS it has also beenrecognized that the complex and heterogeneousnature of urban sediments limits the extent towhich these measurements can be used as sedi-ment source tracers. For example, Charlesworth& Lees (2001) consider that given the largenumber of sources of magnetic material in urbanenvironments, and the processes that changemagnetic signatures, unmixing individual com-ponents of the signal becomes very difficult.

The majority of contaminants to RDS arederived from intrinsic sources, and the major

sources recognized are shown in Table 6.1. Asstated previously, Pb is predominantly derivedfrom leaded fuel (where tetraethyl-lead is usedas an additive). Lead levels in sediment havedeclined, however, with the widespread reduc-tion in use of leaded fuel (e.g. Nageotte & Day1998). Copper and zinc have both been sourcedto vehicle activity, with Cu coming from cor-roded car bodywork (Beckwith et al. 1986) and Zn and Cd being derived from tyre wear(Hopke et al. 1980). Chromium, bromine andmanganese are also present in tyres and brake

leaded petrol (Nageotte & Day 1998; Massadeh & Snook 2002). Clear spatial distributions of metal concentrations are present. For example, lead concentrations are higher in sedimentsin inner city sites, than from those in outer city sites (Case Fig. 6.1a), reflecting the contrastinglevels of vehicular activity. Analysis of the spatial distribution of RDS composition shows thatmetal levels are heterogeneous across the city, with different hotspot locations for differentmetals. This illustrates the localized nature of contaminant inputs into RDS. Mineral magneticanalysis of Manchester RDS (Robertson et al. 2003) reveals that the magnetic fraction is com-posed predominantly of ferromagnetic multidomain particles, indicating inputs of predominantlyanthropogenic origin, derived primarily from automobiles.

Petrographic and mineralogical analysis shows that the sediments, in addition to quartz andclay grains, comprise a significant component of anthropogenic grains (Case Fig. 6.1b). These areeither iron-rich combustion products (mostly iron oxides) or iron-rich glasses, probably alsoderived from vehicular combustion. These grains contain high levels of heavy metals (e.g. Pb andCu up to 1500 ppm, and Zn up to 10,000 ppm) and are probably the major hosts for metals inthe Manchester RDS. Sequential chemical extraction schemes (see Chapter 1) show that mostmetals are bound up with the reducible fraction (probably iron oxides and glasses), but in thecase of Zn a significant fraction is bound in the much more environmentally available exchange-able fraction (Case Fig. 6.1c). This suggests that Zn is likely to be the most mobile metal in the RDS in Manchester and release into road runoff may take place; an observation consistentwith storm runoff analysis. The RDS also contains significant levels of soluble nitrate, chloride,sulphate and phosphate. These anions are key components in river water, of which nitrate andphosphate are key nutrients. It is, therefore, likely that these sediments represent an input ofdissolved species into urban rivers, with subsequent impacts upon water quality.

Relevant reading

Massadeh, A.M. & Snook, R.D. (2002) Determination of Pb and Cd in road dusts over the period in which Pbwas removed from petrol in the UK. Journal of Environmental Monitoring 4, 567–72.

Nageotte, S.M. & Day, J.P. (1998) Lead concentrations and isotope ratios in street dust determined by electrothermal atomic absorption spectrometry and inductively coupled plasma mass spectrometry. Analyst123, 59–62.

Robertson, D.J., Taylor, K.G. & Hoon, S.R. (2003) Geochemical and mineral characterisation of urban sediment particulates, Manchester. UK. Applied Geochemistry 18, 269–82.

ESC06 9/9/06 3:01 PM Page 195

196 KEVIN TAYLOR

linings. Multi-element analysis of RDS, coupledwith principal component analysis, has beenused in some studies to aid in elemental sourceapportionment. De Miguel et al. (1997), in astudy of Madrid, recognized distinct groups ofelements. Those derived from vehicles and con-struction sources were Br, Cd, Co, Cu, Mg, Pb,Sb, Ti and Zn, whereas those deriving from natural soil material were Al, Ga, La, Mn, Na,Sr, Tl and Y.

The platinum group elements (PGEs) Pt, Pd andRh are a relatively recent contribution to RDS,having been emitted into urban environmentssince the early 1990s. The PGEs act as catalystsin catalytic converters and, with the phasing-outof leaded fuel, are currently the metals of mostconcern emitted from vehicle exhausts. There isevidence from many studies for the widespreaddispersion and accumulation of PGEs in RDS,as well as urban soils and airborne particulates.Concentrations of PGEs significantly above thoseof average upper crust values have been reportedfor RDS for cities in Europe and Australia (Wei& Morrison 1994a; Motelica-Heino et al. 2001;Whitely & Murray 2003). Although PGEs inmetallic form are generally considered to be bio-logically inert, soluble PGE salts are indicated tobe much more bioreactive (Farago et al. 1998)and so the presence of PGE in RDS is potentiallyof significant concern.

There is a whole suite of organic contaminants(so-called persistent organic pollutants) sourcedto RDS. These include PAHs (polyaromatichydrocarbons), PCBs (polychlorinated biphenols),hydrocarbons, dioxins, pesticides and herbicides.The sources of these are various and includeboth atmospheric and land-based sources. Forexample, PAHs were observed to be sourcedfrom biomass burning and vehicular emissionsin Vancouver, Canada (Yunker et al. 2002). Prob-ably the largest source of organic pollutants are those derived from vehicular activity. Many of these are found in petrol or diesel (includ-ing benzene, toluene, naphthalene, PAHs), orassociated with automobiles (including ethylene glycol, hydraulic fluids, styrene, oil lubricants).Pesticides and herbicides are applied directly topavements or to urban soils in residential areasand gardens, where they can be removed fromrunoff or erosion and deposited in RDS.

6.2.1.2 River, canal, dock and lake sediments

The range of sediment sources for rivers andcanals is greater than that for RDS, in that aswell as the input of RDS into river sediments,upstream and downstream input of channel-associated material is a major contributor tothese urban aquatic sediments. Although theorigins of sediment in river basins have been

Table 6.1 Contaminant sources to road-deposited sediment.

Contaminant Source

Pb Petrol combustion, paint, smelters, coal combustionZn, Cd Tyre wear, galvanized roofs, abrasion of vehicles, lubricating oils, alloysCu Brake linings, alloys, metal industryFe Car exhaust particulates, corrosion of vehicle body work, background geologyMn Tyre wear, brake linings, background geologyCr Engine wear, vehicle plating and alloys, road surface wearNi Engine wear, metal industry, background geologyAsbestos Break clutch liningsCl, Na Road saltPGEs (Pt, Pd, Os) Catalytic convertorsPesticides/herbicides Garden applicationPAHs Biomas burning, petroleum combustionPCBs Petroleum combustion, industryBacteria Sewage treatment works, animal faecesPharmaceutical compounds Sewage treatment works

ESC06 9/9/06 3:01 PM Page 196


studied extensively (see Chapter 3), very limitedstudy has been made of river sediment sources in urban catchments. A recent exception is that of Carter et al. (2003) for the Aire–Calder riverbasin in eastern England (see Case Study 6.2).Through the application of well-developed stat-istical fingerprinting techniques (see Collins &Walling 2002) they showed that the sediment

in the urban river sections was sourced fromchannel bank erosion (18–33%), uncultivatedtopsoil (4–7%), cultivated topsoil (20–45%),road-deposited sediment (19–22%) and sewageinput (14–18%). The high contribution of urbansources (up to 40% sewage and RDS) illustratesthe marked contrast of urban sediments to thosein non-urbanized catchments. This is in general

Case study 6.2 Sources and dynamics of urban river sediments: Rivers Aire and Calder, West Yorkshire, UK

Limited information is available on the specific sources of sediment to urban rivers, in directcontrast to that of other river systems (Chapter 3). Exceptions are the Rivers Aire and Calder in West Yorkshire (eastern England; Case Fig. 6.2a), which has seen recent detailed analysisthrough research programmes directed at understanding fluvial and urban environments (The Land–Ocean Interaction Study (LOIS) and the Urban Regeneration in the Environment(URGENT) programmes of the UK Natural Environment Research Council). These rivers arepart of a catchment that in its upper parts is predominantly rural in nature, but which in itslower parts runs through heavily urbanized catchments.

Sediment sources in the Aire and Calder have been shown to vary between the upstreamrural-dominated and the downstream urban-dominated parts (Carter et al. 2003). Suspendedsediment in the upper reaches is predominantly sourced from channel-bank material (43–84%)and from uncultivated topsoil (16–57%). This is in contrast to the lower, urbanized reaches,where road-deposited sediment (19–22%) and sewage treatment works (14–18%) acted assignificant sources of suspended sediment (see Case Fig. 6.2a). Carter et al. (2003) also found thatthe proportions of sediment sources to the urbanized reaches varied over individual storm events.Road-deposited sediment sources were highest at the end of high-flow events (Case Fig. 6.2b) as a result of the time lag to flush sediment from road surfaces.

The quality of urban river water is commonly poor and the role of sediments, and sediment-borne contaminants, on water quality and contaminant fluxes is poorly understood. Goodwinet al. (2003) and Old et al. (2003) have quantified the nature of suspended sediment fluxesthrough the Bradford Beck, a tributary of the River Aire, which runs through the heavily urban-ized city of Bradford. Goodwin et al. (2003) carried out continuous monitoring of the urbansegments of the river, both for discharge and suspended sediment loads. The hydrographs show that the river responds rapidly to rainfall events and suspended sediment concentrationsare also high during these events (up to a maximum of 1200 mg L−1). They proposed that in the urban segments sediment was derived from road runoff and combined sewer outflows. Ascan be seen from Case Fig. 6.2c, suspended sediment loads were several orders of magnitudehigher during storm events than during low flow times. Although Goodwin et al. (2003) do not publish compositional data for suspended sediments, as well as the physical impact of the suspended sediment, the flux of contaminants on these sediments is likely to be highlysignificant also.

For the same river, Old et al. (2003) studied the sediment dynamics for a single, large con-vectional summer rainfall event (Case Fig. 6.2c). They found that during this event, over a

ESC06 9/9/06 3:01 PM Page 197

198 KEVIN TAYLOR

Case Fig. 6.2 (b) Variation in sediment sources to suspended sediment in the urbanized portion of the Aire–Calder over asingle high-discharge event. (From Carter et al. 2003.)

Case Fig. 6.2 (a) Location of the Aire–Calder catchment, eastern UK. Also shown are the relative source contributions to thesuspended sediment in the catchment in both upstream and downstream sections. Note the significant component of urban-derived sediment (road-deposited sediment and sewage treatment works (STW) solids) in the sections downstream of the majorurban conurbations. (Data from Carter et al. 2003.)

10 km

N

Leeds

Halifax

Huddersfield Dewsbury

Keighly

Kildwick

Bell Busk

Otterburn

(a)

Beal

Methley

KeyChannel bank

TopsoilUncultivatedTopsoilCultivated

STW solidsRoad-depositedsediment

(b)

12

3

45

Dis

char

ge (

m3 s

−1)

Time since start of discharge rise (hours)

Key

Channel bank

TopsoilUncultivatedTopsoilCultivated

STW solids

Road-depositedsediment

1 2 3 4 5

Rel

ativ

e co

ntrib

utio

n (%

)

0

20

40

60

80

100

10 20 30 40

100

200

300

ESC06 9/9/06 3:01 PM Page 198


period of 15 minutes, water discharge increased from 0.45 to 34.6 m3 s−1. During the sameperiod suspended sediment concentrations reached a peak of 1360 mg L−1. At the peak of thissediment discharge event the sediment flux reached 47 kg s−1. Old et al. (2003) concluded thatalthough the Bradford catchment represents only 3% of the catchment area of the River Aire, it can at times be a major contributor of fine sediment. Given the high contamination of suchfine sediments, storm events are likely to have major impacts upon river quality downstream of urban areas.

Relevant reading

Carter, J., Owens, P.N., Walling, D.E., et al. (2003) Fingerprinting suspended sediment sources in a large urbanriver system. The Science of the Total Environment 314–316, 513–34.

Goodwin, T.H., Young, A.R., Holmes, M.G.R., et al. (2003) The temporal and spatial variability of sedimenttransport and yields within the Bradford Beck catchment, West Yorkshire. The Science of the Total Envir-onment 314–316, 475–94.

Old, G.H., Leeks, G.J.L., Packman, J.C., et al. (2003) The impact of a convectional summer rainfall event onriver flow and fine sediment transport in a highly urbanised catchment: Bradford, West Yorkshire. The Scienceof the Total Environment 314–316, 495–512.

Case Fig. 6.2 (c) Discharge, suspended sediment concentrations (SSC) and suspended sediment load (SSL) for an urbanizedportion of the Bradford Beck, Aire–Calder catchment, eastern England. Note the high levels of SSC and SSL associated with thissingle high-flow event. (Modified from Goodwin et al. 2003.)

0 0.25 0.5 0.75 1.0 1.25

0

1

2

Elapsed time (days)

Dis

char

ge (

m3

s−1)

0 0.25 0.5 0.75 1.0 1.25

0

400

600

Elapsed time (days)

Sus

pend

ed s

edim

ent

conc

entr

atio

n (m

g L−1

)

0 0.25 0.5 0.75 1.0 1.25

0

500

1000

Elapsed time (days)

Sus

pend

ed s

edim

ent

load

(kg

15

min

−1)

200

(i)

(c)

(ii)

(iii)

ESC06 9/9/06 3:01 PM Page 199

200 KEVIN TAYLOR

agreement with Nelson & Booth (2002), whofound that as well as landslides and channel bankerosion, 15% of sediment in an urbanized catch-ment was from road surface erosion. Furthersuch studies are required before a full appre-ciation of the relative contribution of sediment and contaminant sources to urban rivers can begained. As well as contaminant input from roadrunoff, increased levels of nutrients (especiallyphosphorus) and micro-organic pollutants (e.g.pharmaceutical products) are sourced fromsewage treatment works (Owens & Walling2002; Warren et al. 2003). Industrial processesalso source metal contaminants to urban rivers(Walling et al. 2003).

Compared with rivers, which receive sedimentfrom a wide area, canal sediment is commonlydominated by material that is more locallyderived, as a result of the limited transport ofsediment in canals. This may be derived fromindustrial sources or sewage, as well as naturalmaterial eroded from nearby land and road sur-faces. Major canal and dock systems that havesignificant water inputs from rivers, however,can have a significant sediment source from out-side the system. For example, Qu & Kelderman(2001) showed that sediment, and associated con-taminants, in the Delft canals, The Netherlands,have been derived predominantly from the RiverRhine, with the remainder coming from urban andindustrial sources. Urban docks and canals alsocommonly receive high levels of organic matter,discharged from combined sewer overflows,and contaminants derived from boat traffic, forexample hydrocarbons and tributyl tin (e.g.Wetzel & Van Vleet 2003). Within urban lakes,sediment sources are generally a combination of both eroded soil material from the surround-ing catchment and anthropogenic material fromthe urban environment. Atmospheric depositionmay also be an important source of particulatesand associated contaminants, especially for lakeswith no direct river input (Charlesworth &Foster 1999). Sediments deposited within lakesystems are probably highly catchment specificand observations made cannot readily be appliedto other urban catchments (Charlesworth &Foster 1999).

6.2.2 Sediment transport processes

The transport of sediment in urban environmentsis complex. There is a relatively poor under-standing of the pathways that sediments takefrom their source to receiving water bodies, the rate of sediment transport, the location ofshort-term and long-term sinks, and how thesepathways have an impact upon the longer termfate and distribution of contaminants in theurban environment. The movement of sedimentthrough the urban environment can be repres-ented in what has been termed the urban sedi-ment cascade (e.g. Charlesworth & Lees 1999;Fig. 6.4). This cascade recognizes the relationshipbetween sediment sources, transport mechan-isms and deposition (storage) of sediment. Theurban sediment cascade is a highly dynamic sys-tem. As in many depositional systems there arestages of temporary deposition, or storage, priorto further transport. The top of the cascade isconsidered to be the sources of RDS. The deposi-tion of sediment upon street surfaces is highlytransient in nature, with remobilization takingplace down the cascade.

The bulk of sediment transport in the urbanenvironment takes place through the action ofwater, but local redistribution of sediment uponstreet surfaces may also take place by wind.Storm drains carry urban runoff from the streetsurfaces (as well as other impervious surfaces)to receiving water bodies (rivers, canals or docks;Fig. 6.5). In addition to this surface water runoff,the urban drainage system also has to deal withindustrial and domestic wastewaters and sewage.In general, there are two main types of urbandrainage: combined systems and separate systems.In combined systems road runoff is transferredto a common sewer, and from then onwards to asewage treatment works. In this case, these com-bined sewers rarely have sufficient hydrauliccapacity during storm events, and the water isdischarged directly (including sewage) througha combined sewage overflow outlet (CSO) intoreceiving water bodies (Fig. 6.5). In more recentlybuilt drainage systems, a separate sewer sys-tem transports runoff directly to receiving waterbodies, without involvement in the domestic

ESC06 9/9/06 3:01 PM Page 200


sewage system. Once within rivers, sedimenttransport takes place via channel transport pro-cesses typical for rivers in general (see Chapter 3).The temporal and volumetric scales of this pro-cess, however, contrast with rivers from non-urbanized environments (see section 6.3).

Street surfaces are subaerial and during dryperiods resuspension of fine sediment particulatesby wind can readily take place. The suspensionof sediment particulates has major impacts on

urban air quality. Miguel et al. (1999) attributed5–10% of the allergenicity of atmospheric sus-pended particle matter in California to road dustemissions. Particles less than 10 μm (PM10) andless than 2.5 μm (PM2.5) have been measuredextensively in urban air, and a major componentof these particles, especially those above 2.5 μmin size, can be derived from road-deposited sedi-ments (e.g. Hosiokangas et al. 1999; Harrison et al. 2001).

Contaminant sources Sediment sources

Mixing

Deposition

Accumulation

Road-deposited sediment(RDS)

Differentialtransport byparticle size

Removal of fines

Furthermixing

Changingenvironmental

conditions

Chemicalchanges

Attrition

Physicalchanges

Remains asRDS

Deposition ingully pots

Removal viastorm sewers

Chemical changes

Cleaning and removal from system

Physical and chemical changes duringtransport

Input intorivers

Floodplainaccumulation

Transport assuspended sediment

Deposition in canals,docks or lakes.

Growth of authgenicminerals (e.g. phosphates,

carbonates, sulfides).Sink for contaminants.

Dissolution of grains and Fe/Mn oxides

Contaminant release to water column

Early diagenesis

Fig. 6.4 The urban sediment cascade,showing the pathways of, and changes in,urban sediment particulates from sources to sinks. (Modified and extended fromCharlesworth & Lees 1999.)

ESC06 9/9/06 3:01 PM Page 201

202 KEVIN TAYLOR

In addition to street surfaces, the major sitesof sediment deposition in the urban sedimentcascade are gully pots and storm sewers, rivers,canals and docks, and lakes. Deposition (storage)on street surfaces, gully pots and storm sewers is short-term, as is the storage of sediment inurban rivers. In all of these systems sediment ismoved during high-flow events associated withrainfall. Sediment storage in canals, docks andlakes is longer term and removal is generally byengineering activities (e.g. dredging). These areoften termed receiving water bodies and sedi-ment accumulation in these systems can lead to long-term physical and chemical impactsupon water and ecology, as well as providing ahistorical record of urbanization.

As well as physical transport down the urbancascade, chemical changes may take place in thesediment from source to final deposition. Thesetake place as a result of chemical and biochem-ical interactions between water and sediment,including weathering, adsorption, desorption,mineralogical transformations and diagenesis. Astudy by Charlesworth & Lees (1999) concludedthat metal levels in sediments changed duringtransport from street surfaces, through gully potsand sewers and into longer term sinks (rivers,lakes). Some elements (e.g. Cu) increased in con-centration during transport through the cascade,probably as a result of their strong affinity forsolids, and the dominantly fine-grained natureof transported sediments. In contrast, some ele-ments (e.g. Cd) decreased in concentration in

deposited sediments, probably reflecting theirgreater affinity for the aqueous phase.

In contrast to other sedimentary environ-ments there has been little attempt to determinesediment budgets in urban sedimentary environ-ments or model sediment movement. Exceptionsare small-scale modelling of sediment transportin sewer systems from an engineering perspec-tive (section 6.3.2), and temporal measures ofsuspended sediment transport in urban rivers(section 6.3.3). In particular, transport pathwaysof contaminants in urban sedimentary environ-ments are complex and much work is needed to fully understand the longer term impacts ofsediment-borne contaminants to atmosphericand aquatic systems. Furthermore, site specificitylimits generalizations that can be made betweendifferent urban catchments.

6.3 SEDIMENT ACCUMULATION

6.3.1 Road-deposited sediment

Road-deposited sediments have probably receivedmore attention than other urban sediments inrecent years as a consequence of their potentialimpact upon urban air quality and urban run-off. They are also easy to sample and have thepotential to act as a very good proxy for urbanpollution levels. Road-deposited sediment is com-posed of a wide range of sediment grains, whichare dominated by quartz, clay and carbonates.

Sediment inputs - see Fig. 6.3

To sewagetreatmentworks

Accumulationin gully pot

Sewer pipe

Surface runoff

Flowdown drains

Stormoverflow

Suspendedsediment

Fig. 6.5 Schematic diagram of thepathways of sediment transport inthe urban environment.

ESC06 9/9/06 3:01 PM Page 202


In addition, abundant anthropogenic grains arepresent, including glass particles from industrialprocesses and high temperature combustion,metal slags, cement grains, metallic fragmentsand iron oxide particles (Fig. 6.6).

Road-deposited sediments characteristicallyexhibit high concentrations of metals, an aspectof which most research has been directed towards(e.g. Beckwith et al. 1986; Kim et al. 1998;Charlesworth et al. 2003a; Robertson et al.2003). Levels of metals in these sediments arecommonly an order of magnitude higher thanthose of natural sediments. Table 6.2 shows typical values of metals in RDS for a number of urban environments. There is clearly a widerange of values, and Charlesworth et al. (2003a)analysed these data to show that a significantpositive relationship exists between populationsize and metal levels in RDS. Particular study

has been made on the levels of Pb within urbanstreet sediments, largely based on concern overhuman exposure to toxic levels of this contamin-ant. A significant enrichment in Pb levels in innercity centres has been documented (Duggan &Williams 1977; Thornton et al. 1994; Robertsonet al. 2003), supporting other evidence, particu-larly Pb isotopic data (Zhu et al. 2001), that Pbis derived from petrol combustion in such sedi-ments. More recent studies of RDS has shown adecrease in the levels of Pb, consistent with thereduced use of leaded petrol. In 1975 average leadlevels were found to be 941 ppm in ManchesterCity, against a background level of 85 ppm(Nageotte & Day 1998). By 1997 this had fallento 569 ppm (Nageotte & Day 1998). A morerecent study (Robertson et al. 2003) has showna further reduction to an average of 265 ppm in 2000 (see Case Study 6.1). These relatively

(a) (b)

(c) (d)

Fig. 6.6 Sediment grains within road-deposited sediments (backscattered electron images). (a) Spherical Ca–Si–Al–Fe glass grain. (b) Iron oxide fragment, probably derived through the oxidation of a metal Fe grain. (c) Metal-rich glass grain, probably derived fromsmelting processes. (d) Aggregated grain of quartz and clay particles; probably building material. All scale bars = 100 μm.

ESC06 9/9/06 3:01 PM Page 203

204 KEVIN TAYLOR

recent drops in sediment-Pb levels illustrate thetransitory, short-term nature of these sedimentswithin urban systems. Indeed, Allott et al. (1990),using radiocaesium from the dated Chernobylfallout event, documented the residence time of sediment on street surfaces to be short, in the order of 150 to 250 days. Recently, with theintroduction of catalytic converters, attentionhas been directed towards the levels of platinum(and associated elements) within urban streetsediments (e.g. Wei & Morrsion 1994a). It hasbeen documented that Pt levels are increasing in urban sediments, although health impacts ofthese increasing levels remain largely undeter-mined (Farago et al. 1998).

Data on the chemical speciation of contamin-ants within RDS have provided information onthe mineralogical affinity and potential reactiv-ity of contaminants (Fergusson & Kim 1991;Stone & Marsalek 1996; Charlesworth & Lees1999; Robertson et al. 2003). Charlesworth &Lees (1999) found a low concentration of heavymetals associated with the exchangeable phase,results that have been reproduced by other studies. Hamilton et al. (1984), however, foundCd to be associated with the exchangeable phase,and Robertson et al. (2003) found Zn also todisplay a significant affinity to the exchangeable

fraction. Therefore, RDS may be a significantsource of Cd and Zn to urban runoff. Platinumin urban sediments has been shown to be in aform that may be soluble (Farago et al. 1998)and street sediments in gully pots have also beenshown to be actively mobilizing Pt to the aquaticphase (Wei & Morrison 1994a). The majority ofstudies have found most metals to be associatedprimarily with the reducible (Fe and Mn oxide)fraction. Although this suggests that on streetsurfaces contaminant mobility is generally low,changes in pH and redox as a result of depositionin aquatic sediments or sediment water transportwould possibly release metals back into aquaticenvironments. Copper has been shown to displaya higher affinity to organic matter (Hamilton et al. 1984; Robertson et al. 2003). Charlesworth& Lees (1999) ascribed the preference of metalsfor the organic matter fraction in Coventry tohigh levels of organic matter in the sediments.Much less direct information exists on the roleof individual minerals on contaminant behavi-our in urban sediments. McAlister et al. (2000)documented the stabilization of weddellite (cal-cium oxalate dihydrate), derived from sewage,by interactions with metals in street sedimentsof Brazil. In this case, therefore, RDS acted as asink for oxalate, exposure to which has significant

Table 6.2 Typical metal contents (μg g−1) in road-deposited sediments in selected cities. (Data from Charlesworth et al. (2003a) and,for Manchester 2002, Robertson et al. (2003).)

City Population Cd Cu Ni Pb Zn

New York 16,972,000 8 355 – 2583 1811Seoul 10,627,000 3 101 – 245 296London 9,227,687 2.7–6250 61–512 32–74 413–3030 988–3358Hong Kong 5,448,000 – 92–392 – 208–755 574–2397Madrid 2,909,792 – 188 44 193 476Manchester (1975) 2,578,900 – – – 970 –Manchester (2002) 2,578,900 – 32–283 – 25–645 172–2183Birmingham (1976) 2,329,600 – – – 950–1300 –Birmingham (1987) 2,329,600 – – – 527–791 –Taejon, Korea 2,000,000 – 47–57 – 52–60 172–214Amman 1,272,000 2.5–3.4 69–117 27–33 219–373 –Cincinnati 1,539,000 – 253–1219 – 650–662 –Oslo 758,949 1.4 123 41 180 412Bahrain 549,000 72 – 126 697 152Hamilton 322,352 4.1 129 – 214 645Christchurch 308,200 1 137 0 1091 548Lancaster 136,700 3.7 75 – 1090 260

ESC06 9/9/06 3:01 PM Page 204


potential impact on human health, includingdamage to the kidneys and nervous system. In contrast, Serrano-Belles & Leharne (1997)documented the enhanced release of Pb fromRDS upon the addition of chloride, in the formof salt, probably as a result of the formation ofchloro-lead complexes. There is significant scopefor more detailed mineralogical analysis of urbansediments, and the role individual mineral phasesplay in contaminant uptake and mobility.

There have been only a few studies that havelooked in detail at the grain-size distributions in RDS, and the distribution of contaminantsbetween grain-size fractions. In general, RDSdisplays a wide range in grain sizes. Droppo et al. (1998) reported the mean grain size of RDSin Hamilton, Canada to be 227 μm. Sutherland(2003) found the < 63 μm fraction to dominatethe mass fraction of RDS in Hawaii, accountingfor 38% of the sample. Several studies have alsobeen carried out to determine the contaminantloading on different grain-size fractions, withdiffering results. Biggins & Harrison (1986)showed that the mass-dominant fraction of Pbwas in the 250–500 μm fraction, but that therewas a range from 2 to 30% of the mass load-ing in the < 63 μm fraction. Stone & Marsalek(1996) found similar results for RDS in London,UK. Sutherland (2003) found a much higher Pb loading in the < 63 μm fraction for RDS in Hawaii, with this fraction accounting for an average of 51% of the Pb mass. In a similarmanner to sediments from other environments,the increase in contaminant loading in finer grainsizes is generally believed to be a result of theincreased surface area with decreasing grain size,providing greater surface area for metal sorptionto clay minerals or organic matter. The recogni-tion that contaminant loading is heterogeneouslydistributed is important when considering themanagement and pollution abatement of RDS(see section 6.6.1).

The spatial variability of RDS composition hasbeen studied at a range of scales. Studies of RDShave shown that variability exists across thestreet environment, with different levels of con-taminants being present in gutter samples fromthose in street centres and pavements (Linton

et al. 1980). Studies on city-scale variability haveshown that Pb levels are lower in outer city loca-tions compared with inner city sites, indicatingthe role that traffic plays in the distribution ofthis contaminant (Duggan & Williams 1977;Massadeh & Snook 2002; Robertson et al. 2003).Similar patterns have been observed for urbansoil samples. For example, Madrid et al. (2002)showed that metal concentrations are higher insoils in the old quarter of Seville, rather than itsoutskirts. This was put down to vehicular-sourcedmetals. There has been a paucity of systematicspatial analysis of RDS composition over the scaleof a city. A spatial survey of Manchester, UK in2004 (Fig. 6.7; unpublished data) documented a large range in metal levels, with hotspots ofPb, Cu and Zn. These distribution patterns wererelated to both vehicle density and industry.Another spatial survey (in Birmingham, UK) byCharlesworth et al. (2003a) documented a largerange in metal levels, with hotspots of Pb, Cu andZn (Fig. 6.8). These distribution patterns wererelated to both vehicle density and industry.

There has been very limited study into thetemporal variability of RDS. The limited studiesthat have been published indicate that there is temporal variability, especially in the input ofanthropogenic material. Sodium and chlorinelevels in RDS have been shown to fluctuate, withhigh levels being present in the winter months as a result of road salt application. Of particularimportance is the weather, with contaminantlevels (especially Pb) being highest following a number of dry days. Massadeh & Snook(2002), however, suggest that lower Pb levels in Manchester RDS in the summer is a result of lower traffic densities during the summer.Long-term data sets are limited to those on Pbthat were highlighted in a previous section.

6.3.2 Gully pots and sewer systems

Gully pots and sewers are the key elements of subsurface urban drainage systems and themovement and storage of sediments in these hasa marked impact upon both the physical andchemical aspects of urban drainage. The studyof sediment build-up and pollutant loading in

ESC06 9/9/06 3:01 PM Page 205

206 KEVIN TAYLOR

sewerage systems has been predominantly in thefield of civil engineering and a large literature isavailable (for a good overview see Ashley et al.2000). This chapter will focus only on the keyaspects of this topic.

Gully pots are the first entry point of roadrunoff into the urban drainage network, and aredesigned to trap some of the sediment carried bythe runoff. The design is usually one of a sump ora settling chamber, the entry of which is situated

Fe (µg/g)

Zn (µg/g)

Pb (µg/g)

Cu (µg/g)

ManchesterCity Centre

1 km

Fig. 6.7 Spatial distribution of metals in road-depositedsediment in Manchester, UK (Data courtesy of F. Carraz,unpublished).

ESC06 9/9/06 3:01 PM Page 206


at the kerbside. They are a major feature of urbandrainage networks, with more than 17 millionpresent in the UK alone (Memon & Butler 2002a).The trapping of this sediment is desirable fortwo reasons. First, it minimizes the amount ofsediment that enters into the sewerage systemand, thereby, reduces the problems caused bysediment accumulation in sewers. Second, wheregully pots are emptied frequently, it minimizesthe amount of sediment that is potentially flushedout of the sewer system into rivers and receivingwater bodies. The design and assessment of gullypots has been undertaken primarily in the fieldof civil engineering, where the term sludge is used

to describe the sediment in a pot and liquor torefer to the in-place water.

The processes acting within gully pots arecomplex. During runoff events (wet weatherprocesses), denser particles in the water will settle out under gravity (Fig. 6.8). However, thereis usually a high degree of turbulence within the gully pot, which not only limits the amountof sediment that will settle out, but may alsolead to the erosion and re-suspension of existingsediment in the pot. As well as the physical pro-cesses taking place, major biochemical changescan take place within the gully pot (Fig. 6.8).Most biochemical changes take place during

Road surfaceWater and sediment input

Water and fine sedimentCoarsersedimentsettles out

Water and sediment input

Water and sediment

(a) Low flow conditions

(b) Wet weather conditions (high flow)

(c) Dry weather conditions (no flow)

Resuspension

O2NH4

+PO42−

CH4Oxygen consumed

by sedimentChemical speciesreleased from sedimentinto liquor

Water

Sediment

Fig. 6.8 Schematic diagram of thesediment processes operating within agully pot.

ESC06 9/9/06 3:01 PM Page 207

208 KEVIN TAYLOR

periods between runoff events, and are termed‘dry weather’ processes (Fig. 6.8). Many of theseare similar to those taking place during dia-genesis in receiving water bodies (section 6.2.4),although at a smaller scale. Chemical studies of gully pot liquor have shown that the waterreached negative Eh during dry period events,with major impacts upon redox-sensitive chem-ical species (e.g. Morrison et al. 1988, 1995).The major changes are the consumption of oxygen in liquor, changes in chemical oxygendemand (COD) and ammonium, and the releaseof contaminants and nutrients adsorbed ontoredox-sensitive species into the pot liquor (Wei& Morrison 1994b; Memon & Butler 2002b;Fig. 6.8). Dry-weather processes overall increasethe pollutant level of pot liquor, with subsequentnegative impacts on runoff into water ways during wet-weather processes. The longer theperiod of dry weather the greater these changes(Memon & Butler 2002b).

Gully pots are the entry points into the urbandrainage system, but sedimentation within themore extensive sewer system itself has markedphysical and chemical impacts on urban drain-age through the reduction of capacity, leadingto sewer flooding, and the build-up of pollut-ants. Although many sewer systems date fromthe mid-1800s, there have been few advances in sewer design with respect to sediment move-ment. Attention upon sewer design is directedtowards predicting the movement and accumu-lation of sediment. Using hydrological modelsof sediment movement, predictive models forsediment build-up in sewer networks have beendeveloped, with some limited success (e.g.Gerard & Chocat 1998).

6.3.3 Urban rivers

Rivers are natural features of catchments, andare a major component of urban environments,many of which grew up around existing rivers.During the process of urban development riverswere modified to enable navigation, culverted to minimize flooding, and diverted to allowdevelopment. These changes led to the physicalmodification of rivers, to the extent that urban

rivers commonly possess unique physical andhydrological properties. In addition to this phys-ical modification, urban rivers became the majorvectors of domestic and industrial waste removal.A direct result of this was high pollution levels,through both organic and inorganic contamin-ants. It is only with increased discharge legislation,and investment in sewer networks and treatmentplants that urban rivers are now improving inwater quality in developed nations. As high-lighted in Chapter 3, however, the majority ofthe contaminant load in rivers is associated withthe sediment fraction. The residence time of sedi-ment in rivers is greater than that of water, andin spite of water quality improvement many urbanrivers still possess low sediment quality. Thislegacy of pollution is one of the largest problemsfacing urban catchments.

Sediment within urban rivers is itself consideredto be a major non-point source pollutant, as finesediment causes a range of problems in rivers.An increase in fine sediment has impacts uponriver turbidity, affecting biological processesand ecology. In particular, the impacts of highsuspended sediment concentrations upon fishare marked. Impacts include effects on behavi-our and health of fish through gill damage anddamage to spawning grounds, through filling inspaces between sand grains, and reducing oxygendelivery to the fish eggs (Watts et al. 2003). Inaddition, fine suspended material contains thedominant load of metals, nutrients and other con-taminants. An increase in coarse sediment has a physical impact causing channel aggradation,which may lead to channel volume reductionsand flooding. Urbanization can also decrease theerosion rate of the land surface through its coverof concrete, thereby reducing sediment load inrivers. Furthermore, the changes in the flow prop-erties of the river may lead to the downstreamerosion of in-channel sediment due to increasedflow events. This has major impacts upon theecology and biodiversity of urban rivers.

Many of the processes that operate upon sediments, and their interaction with water, inurban rivers are similar to those for other riversystems, and the reader is directed to Chapter 3for full details. The specific study of heavily

ESC06 9/9/06 3:01 PM Page 208


urbanized rivers, and the physical and chemicalcharacteristics of the sediments therein, has beenlimited. It is only recently with the undertakingof integrated projects on urban catchments (e.g. the Natural Environment Research Councilfunded URGENT and LOIS programmes in the UK) that detailed information has becomeavailable. Two major differences can be recog-nized between urban rivers and those from otherenvironments: the spatial and temporal scales of sediment movement, and the level of con-tamination of this sediment.

Sediment in urban rivers can be considered tobe in two major forms. Channel-bed sediment is stored in the river channel and transportedonly rarely by traction and saltation, thus mov-ing downstream slowly. Suspended sediment is carried downstream in suspension during regular flow, with high suspended sedimenttransport under higher energy flow conditions.The former acts as a major storage of sedimentsand contaminants in urban rivers, but the latteris the most important for sediment and con-taminant flux through the river, especially onshort time-scales.

Studies of urbanized rivers have shown theclear increased levels of contaminants associ-ated with the suspended sediment fraction inurbanized river basins. Walling et al. (2003)showed that metals (Cr, Cu, Pb, Zn) and PCBsincreased significantly in urbanized sections ofthe rivers Aire and Calder (north-east England).For example, Cr in suspended sediment in-creased from around 100 μg g−1 in non-urbansections to around 400 μg g−1 in urbanized sections, and PCBs showed a similar fourfoldincrease. Owens & Walling (2002) documenteda clear increase in sediment-bound phosphorus(a major contributor to river eutrophication) as a result of urbanization in the same rivers.They documented changes in total phosphorusfrom < 2000 μg g−1 in upstream sections to over 7000 μg g−1 in sections downstream of urban-ization. They concluded that this increase repre-sented point inputs of phosphorus from sewagetreatment works and combined sewer overflows.The input of these point sources was further sup-ported by a change in the inorganic phosphorus

to organic phosphorus ratio from < 2 upstreamto > 4 downstream of urban centres. This issignificant in that inorganic phosphorus is morebioavailable than organic phosphorus.

Owing to both the impervious nature of urbanland surfaces, and the engineered and culvertednature of urban rivers, the flow in such riversresponds rapidly to rainfall events. This resultsin a rapid rise and fall in the river level, and the river is said to display a flashy response torainfall. This leads to high rates of fine-sedimenttransport in suspension during these flow events,often several magnitudes more than during low-flow periods. Suspended sediment concentrationsin urban rivers can be very high. Gromaire-Mertz et al. (1999) compiled data showing meansuspended sediment concentration for high-flow events to be in the range 49–498 mg L−1.Gromaire et al. (2001) also reported sewer outletsuspended sediment concentrations in the range152–670 mg L−1, illustrating the importance ofsewer outfalls in urban river sediments. Sedimentyields to rivers in urban catchments, calculatedfrom suspended sediment measurement, are in therange of 93.6–479 t km−2 yr−1 (Goodwin et al.2003). A study of the Bradford Beck, Yorkshire,UK (Old et al. 2003) documented the extremelevels of suspended sediment transport during asingle storm event. Suspended sediment concen-trations increased from 14 to 1360 mg L−1 overa period of 15 minutes. A peak sediment flux of 47.2 kg s−1 was recorded, illustrating the highlevels of sediment that are transported by urbanrivers during high flow, and that it is these short-lived events that dominate sediment movement(Case Study 6.2 and Case Fig. 6.2).

The role of suspended sediment in contamin-ant flux is further indicated in Fig. 6.9 for a smallurban river. At low flow, suspended sedimentconcentrations are low, with suspended sedimentloads of only 20 kg h−1. As a result these low-flowstages contribute only low levels of contaminantflux. At high-flow events suspended sedimentloads of over 12,000 kg h−1 are observed, withresulting high levels of contaminant flux (e.g.over 3 kg h−1 of zinc). High-flow events, there-fore, have a significant impact on contaminantinput into receiving water bodies.

ESC06 9/9/06 3:01 PM Page 209

210 KEVIN TAYLOR

6.3.4 Urban canals, docks and lakes

The major sites for sediment accumulation inurban environments are canals, docks and lakes,and these are often the terminal receiving waterbodies in urban catchments. Canals and dockscontrast with lakes in being artificial and heavilyengineered so possess unique hydrological andsedimentological properties. Transport, especiallyfor industry, was a major component in urbanareas and as such navigable waterways and dockswere built to accommodate this. These are mostcommonly freshwater in nature, but docks incoastal urban environments may be marine, andin the case of Venice, Italy, canals may also bemarine in nature. For the majority of urban canalsand docks, water and sediment are input via sur-face water flow or direct industrial or domesticdischarge, and only occasionally are they feddirectly by rivers and groundwater. Owing to thelow-flow conditions in canals and docks, and thesteep-sided nature of these water bodies, they

are highly depositional in nature, resulting inthe rapid accumulation of sediment. This leadsto the requirement for regular dredging to pre-serve navigable status (see section 6.6.2).

Sediments deposited within urban canals anddocks are predominantly derived from anthro-pogenic sources (section 6.2.1.2), and theseanthropogenic sediments have mineralogical andgeochemical compositions significantly contrast-ing to those of natural sediments. To date, how-ever, very little research has been directed towardsdetailing the mineralogy and geochemistry ofthese sediments. Two aspects of urban canal anddock sediments are apparent: they commonlyhave a high organic matter loading as a result ofhistorical sewage input (Boyd et al. 1999; Tayloret al. 2003), and have high metal loadings as aconsequence of industrial waste and discharges(Bromhead & Beckwith 1994; Kelderman et al.2000; Dodd et al. 2003). This high organic and contaminant loading may lead to signific-ant impacts upon overlying water quality as a

1

0.5

0

Sta

ge (

m)

1

0.5

0

Sta

ge (

m)

Low flow

High flow

Q = 4 m3 s−1

SSC = 14 mg L−1

SSL = 201 kg h−1

CuL = 0.1 kg h−1

PbL = 0.06 kg h−1

ZnL = 0.24 kg h−1

Q = 19 m3 s−1

SSC = 185 mg L−1

SSL = 12,664 kg h−1

CuL = 1.27 kg h−1

PbL = 0.76 kg h−1

ZnL = 3.16 kg h−1

Fig. 6.9 An urban river in low flow and high flow, showing typical stage and suspended sediment relationships (River Medlock,Greater Manchester): Q, discharge; SSC, suspended sediment concentration; SSL, suspended sediment load; CuL, copper load; PbL, lead load; ZnL, zinc load. (Photographs courtesy of J. Coyle.)

ESC06 9/9/06 3:01 PM Page 210


result of post-depositional chemical alteration(section 6.4; Case Study 6.3 and Case Fig. 6.3).

Very few published data are available on themineralogical and geochemical association ofthese metal contaminants (Taylor et al. 2003).Canal and dock sediments undergo regulardredging to maintain water depths, and thismaterial is now classed as controlled waste. Thelittle published information on the contamin-ant geochemistry of these sediments is based onhazard assessment work with application to disposal or dredged material (e.g. Bromhead &Beckwith 1994). For example, Kelderman et al.(2000) showed that for the canals in Delft, TheNetherlands, 95% of inner city canal sedimentswere classed as highly polluted, whereas only33% of sediments in the outer city were highlypolluted. There is a requirement for more detailedresearch into the in situ processes operating onsuch sediments, and their role in contaminantcycling in urban aquatic systems.

6.4 POST-DEPOSITIONAL CHANGES IN URBAN SEDIMENTS

As was stated in section 6.2.2, urban sedi-ments undergo physical and chemical changes at many stages of the urban sediment cascade.The two environments in which sediment potentially undergoes major chemical changes(that are most likely to have an impact uponsediment reactivity and contaminant mobility),however, are gully pots and canals, docks andlakes. The former of these has been briefly dis-

cussed in section 6.3.2. This section will focuson the early diagenetic chemical and physicalchanges taking place in sediments within urbancanals and docks.

Early diagenesis is the sum of the processesoperating upon a sediment after deposition andincludes physical, chemical and biological pro-cesses. The early diagenesis of aquatic sedimentsis dominated by a series of bacterially mediatedredox reactions, which result in the oxidation ofcarbon species (organic matter) and the reduc-tion of an oxidized species (Fig. 6.10). Withinsedimentary environments that have an oxy-genated water column, which includes virtuallyall urban water bodies, the primary reactionupon sediment deposition is aerobic oxidation,whereby O2 dissolved in the water is utilized to oxidize organic matter. As O2 is primarilysourced from the overlying water column, thisoxygen is rapidly used up in the first few milli-metres of sediment. The depth of O2 penetrationinto the sediment depends on organic matter con-tent, sedimentation rate and biological activity.In highly organic systems, such as sewage-contaminated water bodies and lakes (e.g. Boydet al. 1999), the sediment oxygen demand throughaerobic oxidation may be high enough to resultin an anoxic water column, especially underlow-flow conditions (Fig. 6.11). Such low-flowconditions are common in steep-sided canals anddocks, and the high organic matter contents ofthe sediments can result in the rapid consump-tion of oxygen in the water column, leading toserious water quality problems.

Manganesereduction

Iron reduction

Sulphate reduction

Methanogenesis

CH2O + 2MnO2 + H2O 2Mn2+ + HCO3− + 3OH−

CH2O + 2Fe2O3 + 3H2O 4Fe2+ + HCO3− + 7OH−

2CH2O + SO42− HS− + 2HCO3

− + H+

2CH2O + H2O CH4 + HCO3− + H+

Aerobic reaction

An

aero

bic

rea

ctio

ns

Reaction Reactants

Aerobicoxidation

CH2O + O2 HCO3− + H+

Products

Fig. 6.10 Early diagenetic reactions taking place within aquatic urban sediment in docks, canals and lakes. (From Taylor et al. 2003.)

ESC06 9/9/06 3:01 PM Page 211

O2

100%

0S

O42−

SW

I

Fe(

II)M

n(II)

CH

4N

H4+

Met

als (

aq)

PO

42−

NO

3−

SR

Ox

O2

SO

42−N

O3−

SROx

Met

h

FeR

+ M

nR

SW

I

Pre

cipi

tatio

nof

viv

iani

te

Pre

cipi

tatio

nof

sid

erite

(if b

icar

bona

teac

tivity

hig

h en

ough

)

(a)

Co

nsu

mp

tio

n o

f ch

emic

al s

pec

ies

(b)

Pro

du

ctio

n o

f ch

emic

al s

pec

ies

If w

ater

bod

y de

epan

d un

mix

ed, w

ater

colu

mn

anox

ia m

ay d

evel

op

Rap

id r

educ

tion

ofni

trat

eLo

w s

ulph

ate

conc

entr

atio

ns in

fres

hwat

er

Sed

imen

ts a

ct a

s a

sink

for

O2,

SO

42− an

d NO

3−

Me

CH

4N

H4+

Sed

imen

ts a

ct a

s a

sour

ce o

f met

als,

nut

rient

s an

d m

etha

neM

etal

s ad

sorb

edon

to ir

on o

xide

sre

leas

ed d

urin

g F

eR

Fe

leve

ls

buffe

red

by

pres

ence

of

sulfi

de

NH

4+ r

elea

sed

durin

g or

gani

c m

atte

rox

idat

ion

Pho

spha

te r

elea

sed

from

org

anic

mat

ter

brea

kdow

n

Pho

spha

te b

uffe

red

by v

ivia

nite

pr

ecip

itatio

n

Met

hane

gas

prod

uced

by

met

hano

gene

sis

Fig.

6.1

1Id

ealiz

ed su

mm

ary

of th

e re

sults

of e

arly

dia

gene

tic re

actio

ns fo

r sed

imen

t and

por

ewat

ers i

n fre

shw

ater

urb

an w

ater

bod

ies (

dock

s, c

anal

s and

lake

s). T

ypic

al v

alue

s and

dep

ths

deriv

ed fr

om d

ata

publ

ishe

d in

Tay

lor e

t al.

(200

3) a

nd D

odd

et a

l. (2

003)

. Por

ewat

er p

rofil

es a

re id

ealiz

ed a

nd w

ill v

ary

with

var

iatio

ns in

sedi

men

t com

posi

tion

and

accu

mul

atio

n ra

te: O

x,ae

robi

c ox

idat

ion;

SR,

sulp

hate

redu

ctio

n; F

eR, i

ron

redu

ctio

n; M

nR, m

anga

nese

redu

ctio

n; M

eth,

met

hano

gens

is; S

WI,

sedi

men

t–w

ater

inte

rface

.

ESC06 9/9/06 3:01 PM Page 212


Upon the consumption of O2, a series of anerobic bacterial reactions are favoured, utlizingoxygen in species such as NO3

2−, Fe2O3, MnO2,SO4

2− (Figs 6.10 & 6.11). Classically these wereseen as taking place through a sequential set ofreactions, giving rise to a series of diageneticzones (Froelich et al. 1979; Coleman 1985). Suchobservations were based on thermodynamic con-siderations and observations from hemipelagicrecent sediments. Such discrete zones are nowrecognized to be the case for slow sedimentationrate, low productivity environments. In urbanaquatic environments the organic-rich nature ofthe sediment results in many of these reactionstaking place simultaneously. These anaerobicearly diagenetic reactions are many and complex,and a complete description of them is beyondthe scope of this work. The major reactions arenitrate reduction, Mn(IV) reduction, Fe(III) reduc-tion, sulphate reduction and methanogenesis(Fig. 6.11). All of these reactions break downorganic matter and, therefore, lead to an overalldecrease in organic matter content as sedimentsare buried. They also tend to result in the decreasein reactivity of organic matter with depth, havingimplications for dredging remediation.

These early diagenetic reactions have animpact upon the short- and long-term fate ofcontaminants in sediments through two prin-cipal mechanisms: release of contaminants intosediment porewater; and the uptake of con-taminants into authigenic mineral precipitates.Within contaminated sediments, metal contam-inants commonly co-precipitate with iron andmanganese oxides. The chemical reduction ofthese oxides (FeR and MnR) results in the releaseof these adsorbed contaminants to sediment pore-waters (Dodd et al. 2003; Taylor et al. 2003).These contaminants then become free to bemoved into the overlying water column, throughthe process of molecular diffusion (Fig. 6.11).This process of contaminant movement fromsediments into overlying water is commonlytermed a benthic flux. This benthic flux has beenrecognized to be the most significant non-pointsource of pollution to water bodies. This releaseof chemical species into porewater during earlydiagenesis is not restricted to contaminants, it

can also be a major pathway for nutrient releasefrom sediments. One example is that of ammo-nium, which is released in the process of organicmatter oxidation. In sewage-contaminated urbanwater bodies, large amounts of ammonium canbe released into sediment porewaters, and thenoxidized to nitrate in the water column. Gasesmay also be generated from sediment duringearly diagenesis. In freshwater organic-rich sedi-ments, methane gas (CH4) is released from sedi-ments through the reaction of methanogenesis(Fig. 6.11). Methane is a flammable and noxiousgas, and so can have major negative impactsupon water quality in urban canals and docks,both aesthetically and chemically.

The build up of chemical species in sedimentporewater also leads to the precipitation ofauthigenic minerals in the sediment. Withinmarine and brackish sediments the predominantmineral formed in this way is pyrite (FeS2). Pyritehas been observed in canal sediments (Large etal. 2002; Taylor et al. 2003) but the absence ofsulphate in freshwater leads to this being a raremineral in urban sediments. The limited studiesof the diagenesis of urban sediments have shownthe iron phosphate mineral vivianite (Fe3(PO4)2)to be the most common mineral (Fig. 6.12; Doddet al. 2003, Taylor et al. 2003). The importanceof these minerals for contaminant mobility is thatmetals can be taken up by these minerals as theyprecipitate, thereby locking up contaminants in the sediment (Large et al. 2002; Taylor et al.2003).

6.5 TEMPORAL CHANGES IN URBAN SEDIMENTS: NATURAL AND ANTHROPOGENIC IMPACTS

The recent nature of the data on urban environ-ments is a limiting factor on the identification oftemporal changes in response to internal andexternal factors. Indeed, in most urban environ-ments there is an urgent need for the collectionof baseline data to enable future changes to bedetermined. As is evident from other chapters in this book there has been extensive use of sediment records to gain insights into historicalchanges in sediment input, land use, climate and

ESC06 9/9/06 3:01 PM Page 213

214 KEVIN TAYLOR

pollution. Such sediments act as ideal archivesfor environmental change (e.g. salt marshes,floodplains). Many archives from fluvial andestuarine sediments have shown the increasingimpact of urbanization on rivers and estuaries(e.g. Cundy et al. 1997; Walling et al. 2003).Many of these pollution changes are mixed inwith inputs from industrial and mining activities,however, and thereby these studies do not givegood, clear information on past trends of urbansediment quality and quantity. Although sedi-ments do accumulate in urban environments, theengineered and disturbed nature of the urbanenvironment has meant that continuous, undis-turbed sediment records are uncommon. Manyreceiving water bodies have undergone dredgingoperations for shipping or remediation, or aretoo shallow to allow the accumulation of fine-grained sediments. There are, however, a limitednumber of studies that have provided informa-tion on short- to long-term changes in urban

sedimentary processes. Short-term records havecome from road-deposited sediment monitoringprogrammes, whereas longer term records havebeen provided by sediment profiles in urbanlakes and canals.

Although no long-term continuous monitoringdata sets exist for RDS composition, individualstudies over two decades on specific cities can becombined to provide useful data on temporalchanges. One such example is Manchester, UK,where Pb levels have been documented to havefallen (see section 6.3.1). A more recent study(Robertson et al. 2003) has shown a further reduc-tion to an average of 265 ppm in 2000. Similarresults where found by Charlesworth et al.(2003a) for the city of Birmingham, UK, where Pblevels in RDS were documented to have droppedby one-third from 1987 to 2002. Recent analyseshave also shown an increase in platinum groupelements in RDS since the early 1990s (e.g. Wei& Morrison 1994a; Motelica-Heino et al. 2001;

(b)(a)

(c) (d)

10µm

P

V

V

V

Fig. 6.12 Minerals precipitated within canal sediment after deposition (backscattered electron (images). (a–c) Vivianite(Fe3(PO4)2.8H2O) taking the forms of radiating needles and laths. (d) Authigenic framboidal pyrite (FeS2; p). Scale bars are shown for each micrograph.

ESC06 9/9/06 3:01 PM Page 214


Whiteley & Murray 2003). Future monitoringprogrammes on RDS composition are likely tofurther reinforce these temporal trends.

Urban lakes are a promising source of urbansedimentation records. Charlesworth & Foster(1999) showed that good records of urban sediment supply and composition were pre-served in two small urban lakes in Coventry,UK. Sedimentation rates in these lakes hadchanged in response to very localized catchmentprocesses, but both clearly showed an increasein metal accumulation after the mid-1950s. Itwas also observed that Pb inputs had decreasedover time owing to the decrease in use of leadedpetrol.

Sediments from urban canals and docks havealso been shown to preserve a good record oflocalized changes in pollution and remediation.Generally, urban canals are too shallow (< 2 m),

and too frequently disturbed to allow a con-tinuous sediment record to build up. In largercanals and docks, however, where water depthsmay exceed 5 m, in the absence of dredging suchrecords can be preserved. One such example isfrom the Salford Quays in the UK where Tayloret al. (2003) documented the clear preservationof pre- and post-remediation sediments andassociated contaminant levels (Case Study 6.3).In more general terms many studies have shownthat surface sediments in urban canals con-tain lower levels of contaminants than deepersediments, indicating a decrease in contaminantinput into urban sediment in recent times (e.g.Kelderman et al. 2000; Bellucci et al. 2002;Taylor et al. 2003). This is generally put downto the environmental legislation reducing dis-charge and the cleaning up of the urban drainageand sewerage system.

Case study 6.3 Sedimentation in an urban water body: Salford Quays, UK

Docks and canals commonly form the terminal receiving water bodies for urban surface runoff,and as such act as significant sites of sediment accumulation and storage. These sediments cancontain significant levels of past and present contamination and so any physical or chemicalperturbations of the sediment can lead to the release of these contaminants back into the urbanenvironment. The Manchester Ship Canal (MSC) was built in 1895 to allow for direct shippingaccess to the City of Manchester Docks at the eastern end of the Ship Canal (Case Fig. 6.3a). At its eastern end the MSC begins at the confluence of three rivers (Irwell, Medlock and Irk)

Case Fig. 6.3(a) Map showing thelocation of the Salford Quays, GreaterManchester, UK (from Taylor et al.2003).

SALFORD

OLD TRAFFORD

N

River I

rwell

ManchesterShip Canal

SalfordQuays

0 500m

MANCHESTER

ESC06 9/9/06 3:01 PM Page 215

216 KEVIN TAYLOR

Depth (m)

Con

cent

ratio

n (µ

g/g

)

Wat

erC

olum

n

Sed

imen

tC

olum

n

0

015

0,00

00

3000

Mn

Fe

Cu

Zn

080

000

30,0

00

0.1

0.2

0.3

Depth (m)

0

0.1

0.2

0.3

0

0.1

0.2

0.3

0

0.1

0.2

0.3

Con

cent

ratio

n (µ

g/g

)C

once

ntra

tion

(µg

/g)

Con

cent

ratio

n (µ

g/g

)

Pos

t-re

med

iatio

n.

Pre

-re

med

iatio

n.

Pos

t-re

med

iatio

nse

dim

ent –

ric

h in

alg

alm

ater

ial.

Pre

-rem

edia

tion

sedi

men

t – r

ich

inco

ntam

inan

t gra

ins.

Met

al-r

ich

glas

s gr

ains

inpu

t dur

ing

site

cle

an-u

p

Cas

e Fi

g. 6

.3(b

)Se

dim

ent p

rofil

e fro

m th

e Sa

lford

Qua

ys sh

owin

g th

e re

cord

of t

he re

med

iatio

n of

the

wat

er c

olum

n. N

ote

the

pres

ence

of h

igh

met

al c

once

ntra

tions

at

12–1

7cm

dep

th, c

ause

d by

the

pres

ence

of a

nthr

opog

enic

met

al-ri

ch g

lass

gra

ins.

Pre

-rem

edia

tion

sedi

men

t is c

ompo

sed

of m

ixed

-orig

in d

etrit

al g

rain

s (na

tura

l cla

y an

d qu

artz

gra

ins,

met

al-ri

ch a

nthr

opog

enic

gra

ins)

. Pos

t-rem

edia

tion

sedi

men

t is c

ompo

sed

of a

utoc

htho

nous

cla

ys a

nd a

lgal

mat

eria

l der

ived

from

the

over

lyin

g w

ater

col

umn.

(Fro

m d

ata

publ

ishe

d in

Tayl

or e

t al.

2003

.)

ESC06 9/9/06 3:01 PM Page 216


that form the major drainage conduit for the Greater Manchester conurbation, and in the past combined sewer overflows have supplied domestic and industrial sewage, and road runoffto the canal. The Manchester Ship Canal is up to 8 m deep, steep sided and up to 50 m wide.Sediment is carried into the canal in shallow, fast moving rivers. As this river water meets thedeep, slow flowing canal, sediment is deposited at the bottom of the canal, as flow strengths are not high enough to keep it in suspension. Owing to the urban nature of the catchment this sediment is organic-rich and highly polluted by sewage and metals. These contaminatedsediments have caused a range of water quality problems (White et al. 1993). Aerobic bacteriaoxidize organic matter using free oxygen in the water. This process, therefore, uses up oxygenfrom the water column, giving rise to what is termed anoxia – oxygen-depletion in the watercolumn. Methanogenic bacteria also break down organic matter in anoxic conditions, and releasemethane gas (CH4). This methane, a flammable and noxious gas, bubbles up to the water column,lifting up mats of sediment and sewage to the water surface (a process termed sediment rafting).

0

10

20

30

40

50

1.5 3

Mn (mg L−1)

SWI

0

10

20

30

40

50

40 80

Sulphate (mg L−1)

Depth

(cm

)

Sed–waterinterface (SWI)

0

10

20

30

40

50

20 40

Fe (mg L−1)

SWI

3

Nitrate (mg L−1)

SWI

1 2 50Ammonium (mg L−1)

SWI0

10

20

30

40

50

0

10

20

30

40

50

Depth

(cm

)

0.2

Zn (mg L−1)

SWI0

10

20

30

40

50

0.1

Case Fig. 6.3(c) Porewater profiles for the Salford Quays sediments showing the post-depositional consumption of chemicalspecies, and the release of metals into sediment porewater. Note the release of Fe and Mn, probably via a combination of bacterialiron and manganese reduction and glass dissolution. (Modified from Taylor et al. 2003.)

ESC06 9/9/06 3:01 PM Page 217

218 KEVIN TAYLOR

6.6 MANAGEMENT AND REMEDIATION OF

URBAN SEDIMENTS

As discussed throughout this chapter the impactsof urban sediment particulates on the environ-ment are wide-ranging and include: impacts uponwater quality of runoff and receiving water bodies, reduction of capacity in drainage systems,an increase in atmospheric particulate concen-trations and a reduction in depth of navigablewaterways. All this means that the active man-agement of sediments is a key requirement in thesustainable urban environment.

6.6.1 Management of road-deposited sediment

As has been described previously, RDS can havea major impact upon waterways, both throughthe volume of sediment transported from these

environments, and the potential for high levelsof contaminants associated with the sediment. It is, therefore, a key element of urban pollu-tion management that levels and composition ofRDS are both monitored and actively managedto minimize their impact. Management practicesattempt to address these issues in two ways. First,target levels for contaminant concentrations canbe set, above which sediment removal is required.Second, routine physical removal of RDS can beundertaken. Although there have been studieson contaminant levels in RDS (see section 6.3.1)there have been no systematic studies on accept-able levels of contamination in RDS, and noguidelines or criteria exist for impacts of con-taminants on urban runoff and water courses.Therefore, the first of these approaches is notcurrently utilized as a management strategy(Case Study 6.4).

As a consequence of this the water quality of the Salford Quays was improved through increasedmixing of air into the water column (Boult & Rebbeck 1999).

The nature of sedimentation in the Quays has markedly changed in response to this watercolumn clean-up (Case Fig. 6.3b). Pre-remediation sediments are composed of a range of naturaldetrital grains, predominantly quartz and clay, and anthropogenic detrital material dominatedby industrial furnace-derived metal-rich slag grains. Post-remediation sediments are composedof predominantly autochthonous material, including siliceous algal remains and clays. At thetop of the pre-remediation sediments and immediately beneath the post-remediation sedimentsis a layer significantly enriched in furnace-derived slag grains, input into the basin as a result ofsite clearance prior to water-column remediation. These grains contain a high level of metals,resulting in a significantly enhanced metal concentration in the sediments at this depth (CaseFig. 6.3b). Significant porewater peaks in Fe, Mn and Zn in the sediment (Case Fig. 6.3c) aremost probably the result of dissolution of these furnace-derived grains in the sediments, possiblymediated by iron and manganese oxide-reducing bacteria. These species have subsequently diffused into porewater above and below the metal enriched layer, indicating the importantcontrol of diagenesis on the long-term fate of contaminants in urban sediments.

Relevant reading

Boult, S. & Rebbeck, J. (1999) The effects of eight years aeration and isolation from polluting discharges onsewage- and metal-contaminated sediments. Hydrological Processes 13, 531–47.

Taylor, K.G., Boyd, N.A. & Boult, S. (2003) Sediments, porewaters and diagenesis in an urban water body,Salford, UK: impacts of remediation. Hydrological Processes 17, 2049–61.

White, K., Bellinger, E.G., Saul, M., et al. (Eds) (1993) Urban Waterside Regeneration: Problems and Prospects.Ellis Horwood, Chichester.

ESC06 9/9/06 3:01 PM Page 218


Case study 6.4 Study of road-deposited sediment management: Tampa, Florida, USA

The management of road-deposited sediments (RDS) is an important component of urban pollution control. The most effective method of control is the removal of RDS through streetsweeping, and disposal. There have been very few systematic studies of the effectiveness of RDSsweeping on improving urban runoff water quality. An exception is a study undertaken for thecity of Tampa in Florida, reported in Brinkmann & Tobin (2003). Generally, street-sweepingfrequency and policy for individual cities is not regulated by scientific information, and thisstudy set out to address this.

Tampa is a mid-sized city, containing sectors of industry, commerce and medium-density residential housing. To assess the compositional variability of RDS within the city, street-sweeping samples were taken from residential, commercial and industrial areas of the city and analysed. The RDS from industrial areas exhibited the highest levels of zinc, copper andbarium, and was linked to industrial activity (Case Table 6.4). Levels of strontium, nickel,chromium and vanadium were highest in commercial areas (Case Table 6.4) and this waslinked to the increased vehicular activity in these areas. Overall, RDS element levels weremostly considered to be non-harmful, but concern over levels of copper, lead and zinc wereraised. The fine-grained fraction of the RDS was greater in industrial and commercial areasthan in residential areas, and this was linked to increased combustion of fossil fuels and industrial emissions. Organic matter was higher in residential areas, due to increased vegeta-tion, and available phosphorus was also higher in these areas, probably due to increased gardenfertilizer use.

Compositional analysis showed that the RDS sweepings removed from the city were com-posed predominantly of inert sands, cement and organic matter. Combined with a high nutrientcontent it was concluded that in many cases these street sweepings could be recycled as topsoiland in soil amendment. For those RDS containing high metal levels however, such a use was notappropriate. If the sweepings were treated by removing the fine fraction (< 63 μm), however,then metal levels would be reduced enough to allow recycling.

A further aspect of this study was to look at the impacts of street-sweeping removal of RDSon sediment and water contamination levels. Sectors of the city were swept either every twoweeks, weekly, twice weekly, or not swept at all. Three conclusions were reached. First, streetsweeping was most effective at reducing RDS accumulation when carried out on a weekly basis.Lower levels of sediment and contamination were found on frequently swept surfaces. Second,street sweeping does result in a marked reduction in contaminant levels in runoff, with the mostsignificant improvements seen with sweeping on a twice-weekly basis. Finally, maintaining aweekly street sweeping schedule is much more effective at reducing sediment and water pollu-tion levels than other factors, such as weather conditions. The conclusion that can be drawn

Case Table 6.4 Compositional variation in road-deposited sediment in the city of Tampa. (From Brinkmann & Tobin 2003.)

Zone Copper Zinc Lead Mass per cent Organic matter Available (ppm) (ppm) (ppm) < 63 μm (%) phosphorus

Industrial 125 96 65 27.3 5.57 353Commercial 23 79 61 24.0 2.18 346Residential 24 59 72 19.0 6.62 520

ESC06 9/9/06 3:01 PM Page 219

220 KEVIN TAYLOR

The routine physical removal of RDS is car-ried out in many urban environments, both foraesthetic reasons and to limit the impact of RDSon watercourses and sewer systems. Removal is accomplished by mechanical street sweeping.Early street-sweeping procedures were carriedout predominantly for aesthetic reasons, andremoval efficiencies were low (Sartor et al. 1974).Subsequent studies have shown that regularremoval of RDS by street sweeping may lead to asignificant reduction in both sediment contamina-tion levels and contamination of surface runoff(e.g. Sartor & Gaboury 1984; also see CaseStudy 6.4). It can, therefore, be a highly effectivemethod of urban pollution management. It hasbeen shown that street sweeping is most effec-tive at removing pollutants in climates wherelong periods of dry weather lead to pollutantaccumulation (Sartor & Gaboury 1984). Streetsweepers can be based on vacuum systems or on a rotary brush system, and comparisons havebeen made on the effectiveness of each type forRDS removal. Generally, although the rotary typeremoves a greater proportion of RDS from streetsurfaces, vacuum-based models are better atremoving the finer grain fractions (Brinkmann& Tobin 2003). This is an important considera-tion as it has been shown that the fine fractionscontain the highest contaminant loading (seesection 6.3.1). There is, therefore, a trade-offbetween an increased volume of sediment removalor more efficient contaminant removal. In general,street sweeping is much less efficient at remov-ing the finer-grained fractions of RDS than thecoarser-grained fractions. This has implicationsfor pollution management as the finer-grained

fraction generally contains the highest loading ofcontaminants. In a study in Hawaii, Sutherland(2003) showed that street sweepers removedonly 62% of Pb from RDS, primarily as a conse-quence of low efficiencies of fine-grain sedimentremoval (Table 6.3). The resulting waste producedfrom street-sweeping can be reused as groundcover or disposed of on land or to landfill, butthis waste has not been widely assessed for itssuitability for such. The limited studies that havebeen undertaken (e.g. Viklander 1998; Clark etal. 2000; German & Svensson 2002) have con-cluded that the sweepings material, owing tohigh contamination levels, should be treated priorto its reuse or disposal on land.

6.6.2 Management of sediment in gully pots

The management of gully pots, and their associ-ated sediment, forms an important part of urbanwater quality management. Most authoritiesregularly empty and clean gully pots, therebyremoving the sediment from the urban drain-age system. This is commonly carried out tominimize flooding and drainage issues, ratherthan for pollution management reasons. Memon& Butler (2002a) modelled the efficiency of gullypots in urban drainage networks and showedthat gully pots can reduce the suspended sedi-ment content of water entering sewer systems(and ultimately receiving water bodies) by 40%, with even larger reductions being possiblewith improved gully pot design. The model alsoshowed, however, that reduction in pollut-ants (such as ammonium and chemical oxygendemand) was minimal.

from this study is that a regular street-sweeping programme, which frequently removed RDS,should be an important component of a city’s pollution management.

Relevant reading

Brinkmann, R. & Tobin, G.A. (2003) Urban Sediment Removal: The Science, Policy, and Management of StreetSweeping. Kluwer, Dordrecht, 168 pp.

ESC06 9/9/06 3:01 PM Page 220


6.6.3 Stormwater management ponds

Stormwater ponds are designed and engineeredto remove pollutants from stormwater runoff,primarily through the settling out of sedimentsfrom the water column. As the majority of thepollutant loading is associated with the sediment,this leads to an improvement in the quality of the stormwater runoff, which can then be discharged to natural water bodies. Pollutantremoval efficiencies of up to 90% have beenreported for such ponds (e.g. Wu et al. 1996).Removal of the nutrient phosphorus has alsobeen documented, but the release of nitrogen in the form of ammonia from the sediments tothe water column may also take place (Hvitved-Jacobsen et al. 1984). The build-up of sedimentsin these ponds leads to a reduction in volume and,therefore, efficiency. Therefore, sediment needsto be removed periodically (see section 6.6.4),but these sediments are commonly contaminated.For example, Marsalek & Marsalek (1997)determined for a stormwater pond in Ontario,Canada, that sediment metal levels were suchthat the sediment could not be reused or placedin residential landfill without treatment.

6.6.4 Sediment dredging

Dredging of sediment is a management tech-nique used on urban aquatic sediments for twopurposes: to maintain draft in navigable canalsand docks, and to remove contaminated sedimentfrom waterways as part of pollution manage-ment. Commonly the two become interrelated,however, as once sediment is dredged for re-

moval, sediment quality guidelines come intooperation to determine its suitability for disposal.With the exception of inland urban docks, mostsediment dredging from urban watercourses isundertaken to remove contaminated sediments,and to limit their impact upon water quality (for details of dredging issues at coastal ports seeChapter 1). Dredged material is either disposedof through land application (e.g. Chen et al. 2002)or, if contaminated, is disposed of to landfill.

An example of sediment dredging of urbancanals for remediation is that of Birmingham inthe UK (Bromhead & Beckwith 1994). Birming-ham canals were built after 1770 and the banksof the canals were heavily industrialized, by for example metal manufacturing, chemical andengineering works. Inputs and waste dischargefrom these activities led to accumulation ofhighly contaminated sediments, with sedimentshaving mean concentrations of 1.0% Cu, 0.7%Zn, 0.3% Cr and 0.15% Pb (Bromhead &Beckwith 1994). Dredging was undertaken to adepth of 1.5 m below water level and resulted in24,000 m3 of sediment removal. Once removedfrom the watercourses, such dredgings are gen-erally treated as waste, and have to be disposedof within strict guidelines.

6.7 FUTURE ISSUES

The major issue of concern in urban sedimentaryenvironments is that of increased urbanizationand industrialization. In the developed worldthe input of contaminants into urban environ-ments will at least stabilize, if not decrease,

Table 6.3 Removal of lead load in road-deposited sediments by street sweepers in Palolo Valley, Oahu, Hawaii. (From Sutherland 2003.)

Grain size fraction (μm) Mean Pb loading (%) Mean street sweeper efficiency (%) Pb load removal (%)

2000–1000 2.7 ± 1.6 83.7 2.21000–500 10.8 ± 6.2 81.7 8.8500–250 13.4 ± 6.2 79.3 10.6250–125 12.8 ± 6.3 75.0 9.6125–63 9.7 ± 2.4 66.7 6.5< 63 50.6 ± 14.9 48.7 24.7Overall 62.4

ESC06 9/9/06 3:01 PM Page 221

222 KEVIN TAYLOR

in response to ever stricter emission controlsand environmental guidelines. As such, urbansediments are likely to become less importantvectors for contaminant transfer through thesediment cascade. The volume of sediment islikely to remain high, however, unless controlsand guidelines for sediment quantity are intro-duced. In contrast, in the developing world theever increasing rate of urbanization, coupledwith a lower level of environmental control, willmean that sediments in urban environments will continue to have a major impact on con-taminant cycling and surface water quality.

In response to this, there needs to be a muchmore sustainable approach to urban develop-ment, and the consideration of sediments willneed to play a key role in this. Sustainable drain-age systems (SuDS) are increasingly being seenas the best way to manage surface water qualityand quantity (Charlesworth et al. 2003b). Thesesystems are designed to slow down the rate of surface runoff, through the use of permeable

land surfaces (the ‘porous’ city). Currently, thesefocus principally on the hydrological aspects of urban environments, but potential exists forintegration of sediments into SuDS. For example,the use of sediments as buffers for pollution (e.g. in artificial wetlands) is currently being pioneered in SuDS.

Climate change will also have a likely impactupon the hydrology, and hence sediment trans-port, of urban environments. Although numer-ous studies on the impacts of climate change on natural systems have been undertaken, littleconsideration has been given to engineered environments. Changes in climate may alter rain-fall and snowmelt patterns, however, therebyhaving an impact on urban drainage. Semadeni-Davies (2004) modelled the impact of possiblefuture climate change on a cold region and cityand showed that frequency and volume of waste-water flows in an urban environment would bealtered, with implications for drainage systemdesign and management.

ESC06 9/9/06 3:01 PM Page 222

7.1 INTRODUCTION

The focus of this chapter is those coastal environ-ments that are associated with the mouths ofrivers: namely coastal deltas and estuaries. Thesedynamic sedimentary systems occur at the inter-face between terrestrial and marine systems and,consequently, their sediment supply, morphologyand functioning are heavily influenced by bothland-based processes, such as river flow and flood-plain development, and marine processes, such astides and waves. Also significant is the fact thatwith the exception of lakes, these environmentsreceive all land drainage, inherent in which isconsideration of water volumes, sediment loadsand contaminants. This chapter examines theseenvironments within the context of how theyfunction as sedimentary systems, how they areinfluenced by changes in natural processes, andhow they are impacted by anthropogenic activity.Useful additional literature relating to estuarineecology (Adam 1990; McLusky 1989), the func-tioning and physical aspects of estuaries (Dyer1997) and estuarine management (French 1997)are recommended. In terms of deltas, furtherdetails relating to the processes of delta formationcan be found in many coastal geomorphologytextbooks (e.g. Woodroffe 2003), and discussionof processes, because of the similarity betweenestuaries and delta channels, falls within the liter-ature cited above. Tropical mangrove-colonizedestuaries and deltas are referred to in Chapter 9.

7.1.1 Definition of estuaries and deltas

It is useful to discuss what is meant by the terms‘estuaries’ and ‘deltas’, and what the difference

7Deltaic and estuarine environments

Peter French

is between them. Both environments representareas of sediment accumulation at the coast, andboth are linked to the mouths of river systems.The chief difference, however, is that estuaries arefeatures that are formed of marine and terrestrialsediments within river mouths in response totheir flooding by a rise in sea-level. Deltas, onthe other hand, develop seawards of the coast-lines where large volumes of fluvial sediment arecarried seawards, at a rate that exceeds the sea’sability to erode it. Accepting this major distinc-tion, however, there are many similarities, notleast of which is that the processes of tides, wavesand freshwater input that operate in estuariesalso occur in the distributary channels of deltas.Therefore, a delta can be regarded as a coastallandform composed of a series of outlets to thesea that are, in effect, estuaries.

The term estuary originated from the Latinword ‘aestus’ meaning tide (Woodroffe 2003).The most commonly adopted definition was firstused by Cameron & Pritchard (1963), when theyreferred to an estuary as ‘a semi-enclosed coastalbody of water which has a free connection to theopen sea, and within which sea water is meas-urably diluted with fresh water derived from land drainage.’ Although being a useful generaloverview of what an estuary is, this definition is generally regarded as an oversimplification asthere is no reference to the tidal processes thatare fundamental in shaping estuaries. E.C.F. Bird(2000) amends this and defines an estuary as:‘the seawards part of a drowned valley system,subject to tidal fluctuations and the meeting andmixing of fresh water and salt water from thesea, and receiving sediment from its catchmentand from marine sources.’ This is more useful

ESC07 9/7/06 3:19 PM Page 223

224 PETER FRENCH

in that it allows greater understanding of theprocesses that link to form the range of estuarytypes (see section 7.1.2).

The term delta was first used by Herodotus in 450 bc to describe the wedge-shaped land-form, resembling the Greek letter delta (Δ), seen at the mouth of the River Nile. A defini-tion based purely on morphology, however, isnot useful when considering the range of deltashapes that can occur. Barrell (1912) and sub-sequently Bates (1953) altered this definition to include some understanding of riverine pro-cesses, before Wright (1982) combined the salientpoints of the two and defined deltas as ‘subaerialand subaqueous accumulations of river-derivedsediment deposited at the coast when a streamdecelerates by entering and interacting with alarger receiving body of water.’ This highlightsthe distinctiveness of deltas over estuaries, inthat the importance of high sediment volumesand the accretion of sediment out onto the coast are emphasized. In terms of both shapeand process, deltas, like estuaries, can be highlyvariable, dependent on volumes of river waterand sediment, and the strength of marine currents(see section 7.1.2).

7.1.2 Nature and significance of deltas and estuaries

Deltas and estuaries, because of their location atthe interface of rivers and seas, can be regardedas major sinks and stores of sediment, particu-larly terrestrial. In addition, the ability of wavesand tides to rework and shape these sedimentsinto recognizable landforms makes these envir-onments very diverse in terms of morphologyand ecology. The sediment that forms deltas and estuaries is also extremely variable, both interms of mineralogy (reflecting sediment source)and grain size (clays to coarse grits). Consider-ing that clays and silts are small sedimentaryparticles, their deposition in these environmentsof high tidal energy, waves and rapidly flowingriver water, is a considerable achievement. A listof the depositional environments where claysare found would largely involve situations wheresedimentation is possible from a still body of

water (lakes and quiet backwaters). Estuariesand deltas do not readily fit into this category.Here we have a unique and significant charac-teristic of these environments. Estuaries and deltasare areas where fresh and salt water meet. Freshwater typically has a salinity of < 0.5 NaCl,whereas sea water is typically around 35. Thismeans that when the two mix to form brackishwater, the salinity will fall somewhere betweenthese figures, depending on the relative propor-tions of salt and fresh water. The importance of this is that the presence of salt enables clayparticles, through electrostatic attraction, to sticktogether to form large sediment grains. This isdiscussed in section 7.2.2, and is importantbecause in addition to clay particle adhesion, pol-lutants can also be adsorbed, making estuarinesediments potentially rich stores for pollutants(see 7.2.3).

Sediment deposition occurs within estuariesand deltas as a result of the complexities of sediment delivery, current activity and marinereworking. Over time, sediment builds up andthe surface becomes covered by tides for shorterperiods of time, i.e. the flood tide will not floodthe sediment surface until later in the tidal cycle, and will leave it sooner during the ebbtide. This means that ultimately, the sedimentsurface can start to be colonized by vegetation.This is fundamentally important in the futuresurvival of the developing marsh because: first,root systems bind the sediment and help it resisterosion; and second, leaves baffle the sediment-laden waves and encourage more rapid sedimentdeposition, thus further facilitating vertical marshgrowth.

Adam (1990) indicates that the first colon-ization by salt marsh plants generally starts tooccur when inundation frequency falls below c. 500 times a year. However, the actual valuevaries between estuaries. In the Blackwater estuary, for example, colonization does not start until inundation frequency falls below 380times a year (Burd 1995). This initial vegetationis salt tolerant (halophytic), because brackishwater will still flood over it once or twice a day.As sediments build up further and tidal cover is gradually reduced, less salt-tolerant vegetation

ESC07 9/7/06 3:19 PM Page 224

DELTAIC AND ESTUARINE ENVIRONMENTS 225

starts to colonize. Mid-marsh vegetation willcolonize when inundation frequencies fall belowc. 230 times a year, and mature marsh below 100(Adam 1990). Ultimately, in the upper reachesof the higher marsh, where tidal inundation israre, freshwater vegetation will appear.

A further significant aspect of deltas and estuaries is a more negative one. The formationof extensive areas of vegetation has, historic-ally, seen this land being regarded as prime fordevelopment. This development has taken manyforms. Initially large areas of salt marshes weredrained and converted to farmland; such areasin the UK include the Fenland around The Wash(47,000 ha), and bordering major estuaries, suchas the Severn (c. 8000 ha). Both of these areashave an almost continuous history of land claimfrom Romano-British times. In other examples,c. 2000 ha of marshes have been claimed sincethe nineteenth century in the Ribble estuary; andin the Dee estuary, around 6000 ha have beenclaimed since the eighteenth century (Davidsonet al. 1991). More recently, with reduced demandfor agricultural land, marshes around majorestuaries and deltas have become prime sites onwhich to develop ports and marinas. Similarly,industrial growth has regarded such land as cheapand highly desirable. The Thaw estuary in SouthWales, for example, was completely claimed in1850 for construction of a power station, largeareas of the Orwell estuary in Suffolk were lostin the nineteenth and twentieth centuries to construct Felixstow docks, and successive piece-meal land claim since the mid-sixteenth centuryfor naval and commercial port facilities hassignificantly reduced the salt marsh resource ofPortsmouth harbour (Davidson et al. 1991) (seealso section 7.5, Fig. 7.14).

Land claim does not only serve to reduce the areas of vegetation in estuaries and deltas,but also has an impact on morphology and func-tioning. Fundamentally, land claim constrainsan estuary, in that it reduces space for water tooccupy. The volume that the flood tide occupiesis known as the tidal prism. Clearly, as moremarsh is claimed, the space available for theflood-tide water to inundate is reduced, and so the tidal prism becomes constrained into a

smaller space. The amount of water entering the estuary remains the same, however, with theresult that the only way of accommodating it is for the height of the water surface to increase.This has some potentially serious implications.First, it increases the risk of defence overtopping;and second, it means that the remaining marshwill be covered more frequently, for longer periods, and to greater depths, potentially lead-ing to vegetation loss. This situation is, in fact,analogous to that which occurs during sea-level rise (see section 7.3). Significantly, how-ever, some examples of historic land claim arenow proving to be short sighted. When claimed,salt marsh soils dewater and contract, falling inelevation relative to any marsh remaining out-side the line of defence. Hence, anything built onthis land is soon below high water mark. In aworld of increasing sea-level rise, this is now amajor problem.

From the above, it can be seen that estuariesare areas of intense human impact and influence,yet they are also areas where natural processescan be particularly dynamic. This diverse rangeof interests and demands placed on estuaries anddeltas represents a key dichotomy. On the onehand, these sedimentary environments are seenas highly dynamic and ecologically important,yet on the other, they are pollutant and sedimentsinks and subject to a range of human pressures,such as land claim and port development. As aresult, many of the large deltas and estuaries of the world have had to be protected throughdesignation of major conservation status.

7.1.3 The classification of deltas and estuaries

Variations in grain size, sediment supply, fresh-water discharge, tidal range and wave activitywill lead to considerable variation between estuaries and deltas. In estuaries, differencesbetween the amount of sea water entering dur-ing each tidal cycle are critical in classification. At the simplest level, estuaries may be classifiedpurely on the basis of tidal range (the amount by which the water level rises between low andhigh tide). Estuaries with a tidal range less than2 m are known as microtidal, those with a range

ESC07 9/7/06 3:19 PM Page 225

226 PETER FRENCH

of between 2 and 4 m, mesotidal, and thoseabove 4 m, macrotidal (Pethick 1984). Giventhat many estuaries have tidal ranges in excessof 4 m, the earlier classification of Davies (1964)classes macrotidal estuaries as being those witha tidal range between 4 and 6 m, whereas rangesin excess of this are known as hypertidal.Although both systems are in use, it is importantwhen referring to a macrotidal system to makereference to which system is being used. A morecomplicated, but more useful system classifiesestuaries on the basis of the amount of mixingthat occurs between salt water and fresh water.Salt water is denser than fresh water, and so in alow-energy estuary there is a tendency for thefresh water to flow over the salt water, the twobodies of water maintaining their integrity, withthe saltwater component forming a ‘wedge’ below

the fresh (Fig. 7.1a). This is a salt wedge, or well-stratified estuary. A typical profile will showthat with increasing depth, the water will staygenerally fresh, then rapidly increase in salinity.Bearing in mind the importance of salt and fresh-water mixing for the deposition of clays, theonly place where this will occur is where the saltwedge meets the fresh water. Clearly, as the tidecomes in and goes out, this point will move upand down the estuary, but regardless of where itis, it is marked by an area of higher turbiditycaused by flocculated clays settling out (see sec-tion 7.2.2). This is called the turbidity maximum.

As tidal range and energy increase, theincreased turbulence causes greater mixing ofthe fresh- and saltwater bodies. In a partiallymixed estuary, there is still an identifiable salt-and freshwater layer but the boundary between

(a) Well stratified (salt wedge)

Fresh

Fresh

Mixing

Salt

e.g. Mississippi SalinityLow High

(b) Partially mixed

FreshFresh

Brackish

Salt

e.g. Mersey / Thames SalinityLow High

(c) Well mixed

Brackish

SalinityLow High

e.g. Severn

Salt

Salt

Fig. 7.1 The classification of estuariesbased on the mixing of fresh- andsaltwater bodies. (Modified from varioussources.)

ESC07 9/7/06 3:19 PM Page 226


them is more diffuse, and the salt wedge is less pronounced (Fig. 7.1b). With increasingdepth, water will remain fresh, then show amore gradual salinity rise until it becomes trulymarine. The final type is the well-mixed estuaryproduced by large volumes of sea water enteringestuaries with significant river discharge. Becauseof the high energy conditions, water mixes com-pletely and is generally brackish throughout itsdepth, although there will be some slight vari-ation in actual salt concentrations with distanceup the estuary (Fig. 7.1c).

The classification of deltas focuses on the degreeof reworking of sediment by the sea. As previ-ously outlined, deltas form when large volumesof river sediment enter the sea at a rate thatexceeds the sea’s ability to rework it. This doesnot, however, preclude that the sea (tides andwaves) will rework the sediment to some extent.Hence, classification of deltas is generally done

on the basis of the dominance of rivers, waves ortides (Fig. 7.2) (see also Carter (1988), Haslett(2000) or Woodroffe (2003) for examples).However, to complicate this, Wright & Coleman(1971, 1972) have suggested that there is a con-tinuum of delta shapes related to the varyingimportance of tides, waves, and river power andthat the tendency of some classifications to assigna delta to one ‘type’ is problematic because it failsto recognize this continuum. Hence, traditionalterms such as ‘lobate’ (deltas with prominentsediment lobes), ‘cuspate’ (deltas with concaveseaward margins caused by wave shaping) and‘birdsfoot’ (deltas where the bifurcating riverhas extended seawards over the delta) are all wellentrenched in the literature but tend to suggestthat deltas are of one type or the other. Overtime, however, these terms have not been usedconsistently, and have also been misinterpretedbecause of difficulties with field identification.

Sediment input

ModernMississippi

Po

Danube

Yukon

Manakam

Ebro

Nile Orinoco

Niger

BurdekinRhoneKelantan

BrazosSao Francisco Copper

Mekong

Yalu

Colorado

Ganges–BrahmaputraKiang–Langat

Fly

Wavedominated

Fluvialdominated

Tidedominated

Loba

te

Cus

pate

Elo

ngat

e Elongate

Estuarine

Wave energy flux Tidal energy flux

Lafourche(Miss.)

St Bernard(Miss.)

Mississippi

Fly

SaoFrancisco

Fig. 7.2 The morphological classification of deltas based on the influence of river, tidal and wave activity. (Compiled and modified fromvarious sources.)

ESC07 9/7/06 3:19 PM Page 227

228 PETER FRENCH

For example, lobate deltas may have distinct cusp-ate traits facing the prevailing wave fronts, yetnot elsewhere (see Woodroffe (2003, pp. 324–5)for a discussion). The appearance of the term‘compound deltas’ may have helped in this respect,but has tended to become a bucket term, whereanything not conforming to the easily identifi-able is placed. Further complicating the issue isthe fact that many deltas change over time asthey mature and as processes such as river input,continental shelf topography, or marine processeschange. For example, the history of the Mississippidelta shows such a pattern (Coleman 1988) (seealso section 7.3, Fig. 7.11). Initially, sedimentlobes form (lobate trait) and undergo enlarge-ment by seaward progradation. This lobate stageis followed by the development of a system ofdistributary channels (birdsfoot trait), which overtime will switch and develop into a complexbirdsfoot pattern. The final stage is for this activedelta front to be abandoned and for a new lobatefront to form elsewhere in the delta system. Inthe Mississippi, there are six major lobate tobirdsfoot cycles, with a typical periodicity in theorder of 1500 years (Coleman 1988).

Despite this complication, as with estuaries,different forms of delta classification may suitdifferent needs. A shape-based approach, as in theWright and Coleman model, remains convenientfor morphological classification needs (Fig. 7.2).If studying how a delta functions in respect tosedimentation and process, however, other formsof categorization may be better suited. In thisrespect, the system suggested by Bates (1953) isuseful as it is linked to processes, and in particulardensity differences, between the inflowing waterand the relatively still receiving body. Where theinflowing river water is denser than the receivingwater, the system is referred to as hyperpycnal,where they are the same, homopycnal, and wherethe river water is less dense, hypopycnal.

7.2 SEDIMENT SOURCES AND SEDIMENT PROCESSES IN

DELTAS AND ESTUARIES

The source of delta sediments is predominantlyterrestrial, whereas estuaries contain a combina-

tion of terrestrial and marine sediments. Fromother chapters, it is evident that terrestrial- andmarine-derived sediment may originate fromdifferent areas: for example, from upland areasvia rivers (Chapters 2 & 3), from urban runoff(Chapter 6), or from the coast (Chapters 8 & 9)and offshore (Chapter 10) (Fig. 7.3).

7.2.1 Sources of deltaic and estuarine sediments

Figure 7.3 highlights the various potentialinputs to estuaries and deltas. Actually quantify-ing these parameters is notoriously difficultowing to problems of measurement and access-ing the more remote parts of the system, such as the incoming tide. In total, however, thesecomponents comprise the sediment budget.Basically, this is a summation exercise in whichall of the inputs are totalled and offset againstthe losses. The resulting value, if positive, willindicate an estuary or delta with a net sedimentgain, whereas if negative, it will indicate onewith a net sediment loss.

In terms of source, the various inputs iden-tified in Fig. 7.3 can be grouped as exogenic andendogenic, depending on whether they origin-ate outside the system (from the catchment orsea) or within (through reworking of existingdeposits or generated within the estuary throughbiogenic activity). Regardless of source, however,there will be considerable temporal variation in

Waterbody

Urban runoffAgricultural

runoff

Industry/portsand harbours

Atmosphericdust

Subtidaldeposits

Intertidaldeposits

Wetlands

Marinesources

Younger

Older

Fig. 7.3 The range of potential inputs to estuaries and deltas(water body). Note that some of these inputs are one way,whereas others represent a bi-directional exchange.

ESC07 9/7/06 3:19 PM Page 228


the relative contribution of each to the sedi-ment budget. The reworking of material fromwithin an estuary or delta (endogenic) is morelikely during the more stormy winter months,or, in the longer term, during episodes of sea-level rise. In terms of biogenic material, bothplants and animals can make a contribution.Plants are perhaps the greatest component ofthis, adding dead plant material to the marshsurface and representing an important elementof marsh vertical growth (see section 7.6.1),although inputs are seasonally variable. Animalmaterial is more variable. Although, in principal,skeletal material is potentially a large contribu-tor (Frey & Basan 1978), it is argued that inreality actual amounts present in a sedimentsequence are typically small because the durabil-ity of shell material is low, particularly whenacidic conditions develop in the marsh sedimentsequence (Wiedermann 1972). Faecal materialis also an important contributor to organic sedi-ment (Frey & Basan 1978).

Microscopic organisms such as diatoms andalgae fill two roles. First, they contribute postmortem to the volume of organic remains in the sediment; and second, they play a key role in sediment stability. Diatoms and algae secretecarbohydrate-rich exopolymers, which serve to stick fine sediment grains together, thusincreasing their resistance to erosion. Dyer(1998) demonstrates the importance of this process in the general salt marsh context, andUnderwood & Smith (1998) in the Humber(UK), Kornmann & de Deckere (1998) in theDollard (The Netherlands), and Riethmüller etal. (1998) in the Wadden Sea (The Netherlands–Danish Coasts); all demonstrate the importancein particular environments.

Input from outside the system (exogenic) is more likely to occur when soils are bare of vegetation or during periods of catchment land-use change. For example, increased sedimentsupply as a result of forest clearance in the catch-ment of the Mahakan delta, Kalimantan causedrapid delta growth (E.C.F. Bird 2000), and mining operations led to accelerated sedimentsupply and subsequent delta growth in theGeorge River delta in Tasmania (J.F. Bird 2000),

the Pahang delta in Malaysia (E.C.F. Bird 2000)and the now disappeared Fal delta (UK) (Bird1998). Conversely, reduced sediment supplycan lead to reduced marsh or delta growth. Theerosion of the Nile delta caused by the buildingof the Aswan High Dam is perhaps the classicexample of this (see Case Study 7.2). Otherexamples where reduced river flow has caused a reduction in sediment supply include theRhône delta, France, the Dneiper delta, Ukraineand the Barron delta, Australia (E.C.F. Bird2000). Although the majority of the examples ofreduced sediment supply are linked to dams andother water-control measures, other causes alsoexist. River dredging can remove large quantit-ies of sediment from the local sediment budget,the cessation of mining activities can removeartificially high rates of sediment supply, andsediment increases following land disturbancecan effectively ‘run out’. As an example of thelatter, the Argentina River delta, Italy, started to erode following reduced sediment supplyafter having initially experienced acceleratedaccretion as a result of increased soil-derivedsediment inputs from land clearance. The sub-sequent erosion of this soil from the hinterlandand the laying bare of the landscape to bedrockled to a cessation in sediment supply (Bird &Fabbri 1993).

In conclusion, the key factor for ensured stability is that whatever sediment the estuary or delta loses, whether via erosion to the sea orvia decreased inputs, it is balanced by newly de-posited sediment, i.e. a positive sediment budget.In the case of deltas, the continued growth of thedelta, and seaward extension of the delta front,is dependent on the continued delivery of sedi-ment from the hinterland. Similarly, the contin-ued vertical accretion of intertidal flats and saltmarshes is dependent on the accumulation of sedi-ment on its surface to replace that removed bythe tide and to compensate for rising sea-levels.

7.2.2 Controls on sediment accumulation andtransport in deltas and estuaries

The sediment inputs described above need to belinked with mechanisms by which they can be

ESC07 9/7/06 3:19 PM Page 229

230 PETER FRENCH

stored and retained within the sedimentary sys-tem. A sediment grain will settle once its fallvelocity exceeds the ability of the water to keepit in suspension. Hence, it is easier for heaviergrains to settle than for lighter ones. Deltas, forexample, often develop a characteristic strati-

graphy with coarser sediments deposited closein shore, grading to finer with increasing dis-tance from the coast (Fig. 7.4), such as is shownin the Mississippi delta (Scruton 1960). Thisland to sea sediment gradation is also marked by structural properties. The coarser delta

Fluvial sediment input

Estuarine/marineprocesses active

Active delta plain

Relicchannel

MHWM

Abandoneddelta lobe

Marginal salt marsh(see Fig 7.5) or mangroves

(see Chapter 9)

Waves and tide reworkingof delta front

Colonization of emergingland with vegetation

Zone of flow expansion,deceleration & dispersion

Plan view

Cross-sectional view

Potential slumping

of delta front sediments

Historic deltafront

Seaward flow of

sedimentladen river

water Tides

Waves

Currents

Outflow &dispersion

Coa

rsen

ing

upw

ards

Exposed surface– soil & vegetationDelta platform

Delta front/slope

Prodelta

Topset – flat-lyinggravels/sands

Foreset – sands, typically inclined10–25o seawardsBottomset – silts, gentlyinclined or flat

Sedimentaryenvironment

Sediments andsedimentary structures

Fig. 7.4 Hypothetical delta section showing typical morphology, processes and grain size trends: MHWM, mean high-water mark.

ESC07 9/7/06 3:19 PM Page 230


platform tends to be typified by horizontal sediment layers (topsets), the finer sands of thedelta slope by seaward dipping (10° to 25°) layers (foresets), and the finest sediments of theprodelta by horizontal to gently inclined layers(bottomsets) (see Fig. 7.4).

Although it is relatively simple to under-stand how a sand or gravel particle can settle,the settling of clay particles is more complex.Estuarine sediments typically comprise largevolumes of intertidal or subtidal muds (mudflats). Clearly, the finer the grain size, the longera sediment particle will take to settle. Stoke’sLaw, a method of estimating such settling times, states that for fine particles less than 100 μm (silts and clays) the settling velocity of a particle is proportional to the square of thegrain diameter. For coarser particles of over 2 mm diameter (coarse sands and above), thesettling velocity is proportional to the squareroot of the diameter. The implications of this are that fine clay particles in estuaries will settlevery slowly. Pethick (1984) demonstrates thataccording to Stoke’s Law, a 2 μm clay particlewill take 56 days to settle through 0.5 m of stillwater. This is clearly not possible in a naturalenvironment because water is only really still for very short periods. Theoretically, therefore,there is no way a clay particle could ever settle in estuarine or deltaic situations. To explain thisapparent discrepancy a mechanism is requiredto enable fine clay particles to settle throughsignificant depths of moving water.

The answer relates to the fact that estuaries areareas of brackish water. Clay particles behaverather differently from sand or silt grains in that they possess a surface attraction, which isamplified by the presence of only a few parts perthousand of salt in the water. Hence, when clayssuspended in fresh water enter an estuary andmix with saline water, their attraction to eachother increases. As clay particles adhere to oneanother they form agglomerations of particlescalled flocs. The name for this process is floccula-tion (Krank 1973, 1975). As more clay particlesstick together and the flocs increase in size, theireffective settling weight increases and so the flocscan settle much faster than the individual grains.

It is this process that explains why estuaries aresuch active sinks for muddy sediments. Import-antly, this ability of clays to attach to other thingsin the water body is not only restricted to otherclay particles. Clays may also attract to a rangeof contaminants, thus making muddy depositspotentially rich in a range of environmental contaminants (see section 7.4.2).

Although flocculation appears a good explana-tion for the occurrence of mudflats in estuaries,many researchers have identified a major discrep-ancy in that there appears to be too much muddeposited if flocculation was the only mechan-ism. McCave (1970) details studies from theGerman Bight that help explain this. He iden-tified that the bottom of the water column ischaracterized by a viscous layer that moves justabove the mudflat surface. This layer contains notjust recently flocculated material that has sunkinto it from above, but also material brought in by the tide, and reworked from elsewhere inthe estuary. From this viscous sublayer, quasi-continuous sedimentation occurs throughout thetidal cycle (Fig. 7.5a). Thus, deposition of claysneed not be restricted to the head of the saltwedge or to periods of slack water.

Flocculation and quasi-continuous sedimenta-tion, therefore, make the formation of extensivedeposits of mud within a relatively high-energyenvironment possible. There are, however, othercontrols over sediment deposition and reten-tion. Even though clay flocculates and formslarger agglomerations, these will fall throughthe water column only when current velocitiesare low enough. On a tidal cycle, water moves at different rates depending on the stage of the tide (Fig. 7.5b). When the tide is fully outand on the turn, water velocity is at its lowest(theoretically, this has to be zero for a period oftime for the water to stop moving out, and startmoving back in). As the tide floods, it picks upspeed before starting to slow again towards hightide. Similarly on the ebb tide, speeds increase as the tide ebbs, before slowing towards lowtide. Current velocities, therefore, will be at theirlowest at high and low water, and at their highestat some point between (note: owing to distor-tion of the tidal wave and the production of tidal

ESC07 9/7/06 3:19 PM Page 231

(a)

Mud

s &

silts

Silt

s &

sand

s

Sal

t mar

sh r

oots

aid

sedi

men

tbi

ndin

g

Sed

imen

t set

tling

&tr

appi

ng w

ithin

sal

tm

arsh

veg

etat

ion

San

ds Infa

una

burr

owin

g &

bio

turb

atio

n le

ads

tom

ixin

g of

sed

imen

t lay

ers

Gul

lyin

g of

sedi

men

ts d

ueto

ebb

run

-off

from

sal

t mar

shes

MH

WS

TM

HW

M

MLW

M

MLW

ST

Silt

s &

mud

s

Flo

od ti

de tr

ansp

ort

Ebb

tide

tran

spor

t

Sal

t mar

shIn

tert

idal

mud

flat

Hig

h in

tert

idal

flat

Mid

inte

rtid

alfla

tLo

w in

tert

idal

flat

Terr

estr

ial

vege

tatio

nU

pper

mar

sh

Junc

usP

ucci

nelli

aH

alim

ione

Spa

rtin

aS

alic

orni

aEnt

erom

orph

a

Mid

mar

shP

ione

erm

arsh

Alg

ae

Sub

tidal

Silt

s &

sand

s5 10 15 20 25Depth (cm)

Typi

cal d

epth

of

biot

urba

tion

by s

peci

es

Corophium

Macuma

Scrobicularia

Nereis

Arenicola

Cardium

Hydrobia

Dep

ositi

on o

f se

dim

ent f

rom

wat

er c

olum

nP

lant

leav

estr

ap s

edim

ent

Dea

d or

gani

c m

atte

rfr

om p

lant

s in

corp

orat

edin

to s

edim

ent

Bur

row

ing

faun

a ca

usin

g bi

otur

batio

n

Set

tled/

cons

olid

ated

mud

s/si

lts

Flo

ccul

atio

n &

settl

ing

Res

uspe

nsio

nof

flui

d m

ud

Ero

sion

of

bed

mix

ing

&tu

rbul

ance

New

ly d

epos

ited

orpa

rtly

con

solid

ated

bed

.C

onso

lidat

es o

ver

time

Qua

si-c

ontin

uous

de

posi

tion

from

fluid

mud

laye

r

Hig

hly

conc

entr

ated

visc

ous

fluid

mud

laye

r

Fig.

7.5

Proc

esse

s affe

ctin

g th

e se

dim

ent d

epos

ition

in e

stua

rine

and

delta

cha

nnel

s: (a

) Cro

ss-s

ectio

n of

a h

ypot

hetic

al c

hann

el sh

owin

g th

e m

ain

proc

esse

s affe

ctin

g se

dim

ent

depo

sitio

n an

d st

abili

ty: M

HW

M, m

ean

high

-wat

er m

ark;

MH

WST

, mea

n hi

gh-w

ater

sprin

g tid

e; M

LWM

, mea

n lo

w-w

ater

mar

k; M

LWST

, mea

n lo

w-w

ater

sprin

g tid

e. (b

)D

iagr

amm

atic

repr

esen

tatio

n of

cur

rent

vel

ocity

var

iatio

n ov

er a

sym

met

rical

tida

l cyc

le (n

ote:

asy

mm

etric

al ti

des w

ill sk

ew th

is g

raph

one

way

or a

noth

er).

(c) R

esul

ting

inte

rtida

lse

dim

ent d

epos

its. (

Parts

b &

c m

odifi

ed fr

om F

renc

h 19

97, p

. 45.

)

ESC07 9/7/06 3:19 PM Page 232


asymmetry, this point of maximum velocityneed not be at the half way mark). The patternsof sedimentation will reflect this (Fig. 7.5c).Where the current velocities are lowest, clays cansettle out, but when they flow faster, clays are keptin suspension and only the coarser sedimentscan settle. Bearing in mind that this variation incurrent velocity is paralleled by a rising andfalling of the tide, these patterns of depositionare shown spatially by areas of sediment accre-tion, such as subtidal muds (below low water),mid-tide sand flats where velocities keep clays insuspension, and high-tide mud flats where clayscan again settle out.

A further aid to increasing sediment transferfrom the water body to the sediment surface isthe presence of vegetation (Fig. 7.5a). Sedimentsurfaces tend to be relatively smooth in respectof the ability to cause friction with the overlying

water body. Hence, the energy lost by wavesmoving over a mud flat is relatively small. Incontrast, vegetation is rough and provides muchmore of a barrier. Waves running over a veget-ated surface will, therefore, use considerablymore of their energy, which means that they will lose speed and drop more of their sedimentload. This ability for vegetated surfaces (see alsomangroves, Chapter 9) to dampen wave energyis important in estuaries and deltas because itnot only facilitates vertical sediment accumula-tion, but also reduces the wave energy reachingthe landward limit of the marsh (often a sea wall).Work on the North Norfolk (UK) marshes(Moeller et al. 1996; Shi et al. 2000) has shownthat after crossing 180 m of a vegetated marshsurface, waves typically lose up to 80% of theirenergy. Not only does this represent an aide tocoastal defence in the area, but it also means that

High

LowCur

rent

vel

ocity

LT HT

Flood tide

SandMud Mud

Mid-tide sand shoals

Time

Sub

tida

l and

low

inte

rtid

alm

ud b

anks

Hig

h in

tert

idal

mud

flats

and

salt

mar

shes

Peak current velocity

High current velocities –mud kept in suspension

Low current velocities –mud settles out

(c)

High

High

Low

Cur

rent

vel

ocity Flood

Ebb

LT HT LT

(b)

Fig. 7.5 (Continued)

ESC07 9/7/06 3:19 PM Page 233

234 PETER FRENCH

the loss of such energy is coupled by a reducedability to carry sediment, leading to increasedsediment deposition.

Attenuation of wave energy due to baffling by vegetation will lead to the increased deposi-tion of sediment directly onto the marsh surface.However, this need not be uniform across the marsh surface. Given that one of the primemechanisms that increases deposition is the lossof wave energy, the areas of greatest depositionwill occur in the parts of the marsh that createthe greatest dampening effect to waves. Studies inNorth Norfolk (UK) have shown that the rate ofsedimentation decreases with distance from boththe seaward edge and creeks. At Hut Marsh onScolt Head Island, North Norfolk, Stoddart et al.(1989) showed that accretion rates near majorcreeks reached 8 mm yr−1, falling to 2 mm yr−1

away from the creeks (Fig. 7.6). This is becauseas the water starts to flow onto the marsh(whether from the seaward edge or having over-flowed onto the marsh surface following entryvia a creek system), the greatest energy loss will occur over a relatively short period of time(see also Allen 2000). Other studies have alsoquantified this change in sedimentation rate.Richard (1978) has shown that for a marsh inLong Island, New York, vertical accretion ratesare typically in the order of 9 to 37 mm yr−1

along the seaward edge of the marsh, falling tobetween 2 and 4 mm yr−1 at the landward edge.These figures, particularly for the low, pioneermarsh, show a wide discrepancy, and this relatesback to the earlier point that the greatest rates of sedimentation will occur in those parts of the marsh that create the greatest energy loss.Even though the seaward edge of the marsh maybe collectively referred to as low or pioneermarsh, there will still be considerable variationin vegetation height, density and exposure towaves and thus the loss of energy and corres-ponding increases in sedimentation rates willvary spatially. Similar depositional patterns andvariation are shown for marshes along the New Brunswick coast, Bay of Fundy (Chmura et al. 2001), and along the Normandy coast ofFrance, where mean vertical accretion rates fallfrom 5.5 to 4.1 mm yr−1 with distance from themarsh edge (Haslett et al. 2003).

In deltas, the pattern of sedimentation alongthe seaward margins is controlled by the pro-cesses acting as the river discharges into the relatively still body of water (see Fig. 7.4 inset).The classification of deltas by Bates (1953) men-tioned in section 7.1.3 is useful in this context.Where deltas form as rivers enter a freshwaterlake, water densities are the same (homopycnal).Here, mud deposits are rare because there is no

Hut Creek

0.2

0.4

0.6

0 25m

N

Hut Marsh

0.4

0.80.6

TheWash

Norwich

Fig. 7.6 Rates of sediment accretion (in mm yr−1) in relation to creek proximity. Data based on HutMarsh, North Norfolk (UK). (Based on data fromStoddart et al. 1989.)

ESC07 9/7/06 3:19 PM Page 234


increase in salt to drive flocculation (an exceptionhere would be a salt lake). In more typical situ-ations, deltas are likely to involve the discharge of fresh water into saline (hyper or hypopycnal),and so flocculation of fine sediment occurs. Saltwedge conditions (see Fig. 7.1) would be ana-logous to a hypopycnal delta (e.g. Mississippi)and the sediment-laden river water will be carriedfurther seawards across the top of the wedge (cf. Fig. 7.1a), being deposited further from thedelta front. Tides and waves may then reworkthis sediment back onto the delta, or along thecoast. If the delta is hyperpycnal (e.g. YellowRiver delta), the incoming sediment load willsink at the front of the delta, to be redistributedby tides along its seaward edge. A detailed dis-cussion of these relationships can be found inWoodroffe (2003).

The distribution of sediment type and thedepositional morphology of deltas and estuariesare, therefore, controlled by a combination ofriver flow, vegetation coverage, tides and waves.It has been assumed so far that tides are simpleevents that involve the rising and lowering ofsea-level. Although this is essentially the case,further complications vary the importance oftides as morphological agents. Tides can facil-itate the reworking of sediment onto a deltafront, or can control the amount of marine sedi-ment entering or leaving an estuary. Hence, theincoming tide brings sediment inshore, and theoutgoing tide takes it offshore. If the ability to move sediment is the same on both incomingand outgoing tides, the net impact will be nochange in the net sediment budget. Frequently,however, this is not the case. Due to variationin, and shallowing of, local sea bed topography,the incoming tidal wave can become distortedby interaction with the long shore profile, out-rushing river flow, and estuarine/delta channelshape, to produce asymmetrical tides. Theimplications of this are that the time taken bythe flood- and ebb-tide can vary. In a flood-tidedominated setting, the flood-tide takes a lot lesstime to fill the estuary than the ebb-tide takes to empty it. The implications here are import-ant because although the time taken to fill theestuary is less than to empty it, the volume of

water involved is the same. Thus, water has tomove a lot faster on a shorter flood-tide. Theresult of this scenario is that a shorter, higherenergy flood-tide will bring a lot of sedimentinto the estuary, and the slower, lower energyebb-tide will allow a lot of it to settle and beretained in the estuary. The converse of this, i.e. a longer flood-tide and shorter ebb, can alsooccur, thus resuspending much of the sedi-ment brought in by the flood tide and moving it out of the estuary. The former situation willlead to the net gain of sediment in the estuary(positive sediment budget), whereas the latterwill result in a net loss (negative sediment budget). Dyer (1997) indicates that the degreeof asymmetry is determined by the relation-ship between an estuary’s volumetric and tidalcharacteristics. In general, he shows that ebbdominance tends to occur in estuaries that aremicrotidal and hyposynchronous, whereas thosethat are macrotidal and hypersynchronous tendto favour flood dominance. In this context, synchronicity relates to the convergence andfrictional resistance of an estuary. Hypersyn-chronicity occurs when convergence exceedsfriction, and hyposynchronicity when frictionexceeds convergence.

7.2.3 Sources of anthropogenic inputs into deltaicand estuarine sediments

Anthropogenic inputs refer to any substancedelivered to deltas and estuaries that is not ofnatural derivation. This includes contamina-tion and ultimately pollution (see Chapter 1 fordefinitions) in the widest sense, i.e. waste prod-ucts from industrial and urban areas, increasedsediment input as a result of dredging or miningoperations, and agricultural products. Figure 7.3shows the range of sediment sources that canenter estuaries and deltas. All of these sedimentsources, however, can also be linked with con-taminants. What is immediately noticeable isthat the range of potential contaminants is verywide, reflecting the fact that estuaries and deltasare sourced from environments that receive themajority of land-based drainage as well as tide-derived marine inputs.

ESC07 9/7/06 3:19 PM Page 235

236 PETER FRENCH

Contaminant problems in estuaries and deltascan, therefore, be particularly acute because theyact as key sediment sinks (see section 7.2.2). In a similar way, any contaminant that enters thesystem in solid (particulate) form will behave as a sediment, and become deposited in the same way. Such contaminants may derive fromdredging, where subtidal sediment is resuspended;from agricultural activity, where exposed soilsmay get washed into streams and eventuallyestuaries; from industry, where tailings mayenter through discharge pipelines; from ports,where the loading and offloading of ships leadsto spillage; and from mining operations, whereerosion of spoil tips leads to inputs of wastematerials.

Particulate contaminants may pose aestheticproblems (Fig. 7.7), but of greater significancefrom a water and environmental quality aspectare the contaminants that enter an estuary ordelta as part of the water body (e.g. metals,

nutrients). Such contaminants may pass directlyout of the system in solution, or they maybecome adsorbed onto clay particles or organicdebris (Fig. 7.8). Similarly, organic contamina-tion such as oil or sewage will break down usingoxygen in the water body, thus increasing thebiological oxygen demand. The longer a con-taminant is present within the estuary, the greaterthe chance of it being broken down or adsorbedonto a clay particle and incorporated within the sediments. In terms of the latter, this meansthat as clay flocs settle out to form mudflats,they can carry with them significant amounts of contaminants adsorbed to their surface. Thechemistry of this process can be very complex(see Stum (1992) or Andrews et al. (1996) forfurther details), but in essence metals (positivecharge) are transferred from the water body tosediment (negative charge) (Fig. 7.8), and thuscontaminants become associated with particlesurfaces. Other modes of contaminant storage

Fig. 7.7 Deposition of coal dust in the intertidal zone of the Severn estuary (Ogmore Beach, South Wales). This material represents thereworking of contaminants from historic mining activity linked to the South Wales coalfield (French 1990).

ESC07 9/7/06 3:19 PM Page 236


involve chemical combination of metals withsulphides and sulphates, or within mineral lattices(see Chapter 1).

The main issue with metal contaminants is thatthey cannot be readily observed, and thereforeattract little attention. The fact remains though,that many muddy estuaries and deltas containsignificant quantities of metals stored within their

sediments. Figure 7.9 shows a typical examplefrom a salt marsh in the Severn estuary. What is noticeable here is that the concentration ofindividual metals varies with depth. This trendis a historical one, with the levels of metalsdeposited in each layer of sediment reflecting thatin the estuary at the time of deposition. Hence,sediment deposited before 1850 contains rela-tively little contamination, whereas that depositedin the 1950s has significantly higher levels. Sucha trend is typical of many industrial estuariesand can be divided into three zones. The lowerzone (I) represents background contaminantlevels, when the estuary or delta sediments were being deposited in times of no industrial-ization, and reflects levels from natural erosion.The middle zone (II) marks a period of rapidlyincreasing contaminant levels, and represents atime of rapid industrialization and decliningenvironmental quality. In the UK, the start ofthis zone can be linked to the onset of the indus-trial revolution in the mid-nineteenth century.The uppermost zone (III) shows declining levelsof contamination, and links to the cleaning upof the environment, increased environmentalawareness and legislation, and a general declinein heavy industries.

Clay

Zn2+Cu+

Cu+

Cu+

Al3+

Zn2+

Zn2+

Mg2+

Al3+K+

K+

Pt2+

Mg2+ Zn2+

Al3+

Pt2+

Fig. 7.8 Hypothetical representation of the adsorption ofmetals to clay particles. The positively charged metals areattracted to the negatively charged clays. As the clays flocculate,so the metals are laid down with the sediments.

10

20

30

40

50

60

70

80

90

100

110

120

130

140

Dep

th (

cm)

40 50 60 70 80 90 100

110

120

130

Cu (ppm)

III

II

I

100

150

200

250

300

350

Pb (ppm)

300

400

500

600

700

800

900

1000

Zn (ppm)

1958

1850

Age

1989

1960

1936

1900

Fig. 7.9 Metal depth profiles from a salt marsh deposit in the Severn estuary, south-west UK. Note the variation with depth, whichreflects the changing level of contaminants within the estuary over time. I, II and III represent ‘chemozones’. These are distinct zonesthat can be identified in industrialized estuarine sediments. Chemozone I represents static, background levels of the uncontaminated,pre-industrialized estuary; II represents increasing contaminant levels during industrial growth; III represents declining contaminantlevels representing industrial decline and increased emission legislation (see text). (Modified from French 1996.)

ESC07 9/7/06 3:19 PM Page 237

238 PETER FRENCH

When still resident in the water body, or evenafter deposition, these contaminants can have asignificant ecological impact, and thus be regardedas pollutants (see Chapter 1 for definitions). For example, toxicity may cause fatalities andendanger parts of the food chain or, if at sub-lethal levels, may have an impact on the abilityto reproduce. Such issues may occur whether apollutant is in solution within the water body,or within the sediment where it remains bio-available. Furthermore, some species can toler-ate pollutants and store them within their bodytissues, a process known as bioaccumulation orbiomagnification (Clark 1992).

Although the environmental levels of contam-inants/pollutants have declined since the mid-twentieth century (see Fig. 7.9), developments inchemical engineering have resulted in the environ-mental presence of a new suite of compoundsknown as organohalogens. These are generallyused as pesticides or in electrical equipment but,importantly, are unable to be broken down and,unlike metals, can be toxic at very low concen-trations. The organohalogens include such groupsof compounds as the PCBs (polychlorinatedbiphenyls), HCB (hexachlorobenzine), HCH(hexachlorocyclohexane, lindane), and DDT(dichloro-diphenyl-trichloroethane) (McLusky1989). These compounds show the classic trendsof bioaccumulation and biomagnification, withthe greatest impact on the higher predators.More significant for estuarine systems is thatmuch of the world’s waste and surplus applica-tion of these substances has drained into themand, because of the accumulatory nature andstorage potential associated with clays, havebecome stored.

Another significant pollutant in deltas andestuaries, particularly in terms of its visibility, is oil. By their very nature, estuaries and deltaslend themselves to the establishment of ports,harbours and oil refineries. Indeed, some of thelarger deltas (e.g. Mississippi) also contain oilfields of their own. Oil spills are a very emotiveissue and whereas the media may not appear too concerned to discuss metal pollution in salt marsh sediments, they will readily report on a large oil tanker spill. Although an oil

tanker accident may appear more newsworthy,a factory releasing small quantities of oil on a daily basis is likely to be more significant in the long term than one major release. Nelson-Smith (1972) cites one example. A typicalrefinery effluent may contain small traces of oil (c. 10–20 ppm) that are not readily detectablewith the naked eye. If, however, this is linked toa discharge of 455,000 L min−1, then on a dailybasis this equates to 6825 L of oil. Ironically,the major accident which receives the greatestpublicity is somewhat of a rarity. Farmer (1997)reports data from the International TankerOwners Pollution Federation (ITOPF) whichshows that only 12% of marine (i.e. not neces-sarily delta or estuarine) oil is derived fromtanker accidents. The greatest source (37%) isfrom industrial sources and urban runoff, suchas that reported by Nelson-Smith (1972), anddirectly affects estuaries and deltas.

Despite the continued inputs of oil, perhapsthe most remarkable aspect of estuarine anddeltaic environments is their resilience. Althoughthe oiled parts of plants may die, they gener-ally grow again once the oil has broken down.DeLaune et al. (1994) studied the impacts of oilon salt marshes by artificially oiling areas ofSpartina marsh and trying different methods ofcleaning, notably leaving the oiled marsh to becleaned by flushing with sea water, applying adispersant, and the cutting and removal of theoiled growth. They found that after 95 days ofmonitoring, there was no major difference in the plots and hence concluded that the bestcourse of action is to leave marshes to recoverand regrow naturally. Such conclusions have alsobeen supported by Gilfillan et al. (1995), whoshowed that 15 years after the 1978 AmocoCadiz spill on the coast of Brittany, France, theareas of marsh where recovery was most suc-cessful were those that had been left alone. Tealet al. (1992) also showed that 20 years after alarge spill in Buzzard’s Bay, Massachusetts,marsh growth was as good as areas unaffectedby the spill. However, in both cases, regenerationtook up to 15–20 years to achieve, suggestingthat regrowth may initially be at a slower ratethan pre-spill.

ESC07 9/7/06 3:19 PM Page 238


7.3 PROCESSES AND IMPACTS OF NATURAL CHANGE IN

DELTAS AND ESTUARIES

7.3.1 Erosion of intertidal sediment substrates

Deltas and estuaries are natural systems and,therefore, undergo change as the processes con-trolling them change (Table 7.1). Such changescan be simple, such as increased wave activity,which may erode more sediment from the delta front or marsh edge, or changes in the volume of sediment brought down stream tosupply the delta. More subtly, changes in stormfrequency may alter the erosion and accretion

patterns of a salt marsh, or increasing sea-levelmay result in changes to water table condi-tions, and thus the stability of stored pollutants,or the stability of marsh flora. The increase inwave activity caused through increased stormi-ness will result in higher energy conditions at the delta front or marsh edge. This could lead to an increase in net erosion and the loss of delta or marsh sediments over a period of many years. In the Severn estuary (UK), forexample, the salt marshes show a series of accretion and erosion events that can be cor-related with episodes of increased storminess(see Case Study 7.1).

Table 7.1 Types of change and the associated impacts in deltas and estuaries.

Nature of change Impact

Reduced sediment supply Inability of vegetation surfaces to keep pace with sea-level rise. Loss of sedimentsupply to mudflats. Loss of supply to delta front, increasing net marine erosion

Increased sediment supply Infilling of channels, levee breaching in deltas. Burial and mortality of infauna andvegetation

Changes in tidal range Emergence or submergence of vegetated surfaces. Variation in water table.Changes to status/position of salt wedge. Changes in mixing. Increased salinepenetration upstream

Changes in storminess Increased erosion and delta/marsh recession (see Case Study 7.1). Increased defenceovertopping

Increasing sea-level Coastal squeeze, marsh loss, increased water table. Landward movement of saltwedge. Increased tidal penetration up-river

Decreasing sea-level Delta/salt marsh advance. Falling water table. Seaward movement of salt wedge. Down river advance of freshwater habitats

Land claim Increased tidal constriction producing net increased sea-level (see above). Loss offlood areas. Coastal squeeze

Dredging Increased tidal prism with potential net drop in sea-level (see above). Modifiedcurrents and tidal flood/ebb patterns

Increased wave activity Cutting back of delta front or marsh edge. Increased sediment input

Case study 7.1 Accretion and erosion cyclicity in the Severn estuary, UK

The Severn estuary is a large, macrotidal estuary system in south-west England. Its maximumspring tidal range is 14.8 m and extensive freshwater drainage from its catchment means that it is a well-mixed and extremely high-energy estuary. The intertidal morphology is complex in that it reflects a series of individual salt marsh units overlying or banked up against othermarshes. The visual effect is that in many locations marsh surfaces descend, step-like towardsthe present channel. Allen & Rae (1987) first studied these marshes and categorized themstratigraphically. The oldest marsh unit, termed the Wentlooge Formation (Case Fig. 7.1a(i)),comprises a complex series of clays and silts, interspersed with peat layers of varying thickness.

ESC07 9/7/06 3:19 PM Page 239

240 PETER FRENCH

Sea level

Accretion

(i) Wentlooge (c. 2000 yr BP)

(a)

Sea level

Accretion

Seabank

Reclaimed land

(ii) Late Wentlooge (late 17th Century)

Retreat

(iii) Early Rumney (early 18th Century)

Reworkedmaterial

s.l.Accretion

(iv) Mid Rumney (early 19th Century)

s.l.

Retreat

(v) Early Awre (late 19th Century)

s.l.

Accretion

(vi) Late Awre (early 20th Century)

s.l.

Retreat

(vii) Early Northwick (1930)

s.l.

Accretion

(viii) Late Northwick (1960)

(ix) Present day

Seabank

Reclaimed land

SeabankMHWST c. 5 m

Bedrock

River Terrace

WentloogeFormation

RumneyFormation

Awre Formation

NorthwickFormation

s.l.

Reworkedmaterial

ModernmudflatRetreat

(b)

Case Fig. 7.1 (a) The sequential evolution of salt marsh morphology in the Severn estuary(French 1999). (b) Post-early nineteenth centurybrown Rumney Formation sediments overlying the older green Wentlooge sediments (Rumney Great Wharf, South Wales).

ESC07 9/7/06 3:19 PM Page 240


These sediments started accreting around 5000 years ago and represent a complex series of different marsh units overlying each other, although the detail of this is difficult to determine atthe present time owing to limited exposure and almost uniform sediment type between differentmarsh units (Allen & Rae 1987).

Much clearer, however, is the more recent morphology. The Wentlooge Formation continuedaccreting until the end of the seventeenth century (Case Fig. 7.1a(ii)). By the early eighteenthcentury, these marshes were eroding (Case Fig. 7.1a(iii)), remobilizing large volumes of sediment.At this time, the quantities of pollutants reintroduced through this reworking were minimalbecause this period occurred before the industrial revolution and, hence, the polluting of theestuary. Erosion continued through much of the eighteenth century but by the early nineteenthcentury rapid sediment accretion had resumed, leading to the deposition of the Rumney Formation(Allen & Rae 1987) (Case Fig. 7.1a(iv)). These sediments were slightly coarser (silty clays ratherthan clays) than those of the Wentlooge, suggesting a new input of silts to the system. Greatthicknesses of Rumney sediments were deposited, leading to the burial of the old Wentloogesurface. This relationship can be seen in many marsh sections, such as those in South Wales(Case Fig. 7.1b). Importantly with regard to the pollution history of the estuary, the accretion ofthe Rumney Formation spanned the industrial revolution, meaning that the levels of pollutantsboth in the estuary and stored in sediments were rapidly increasing (see Fig. 7.9). In addition, theneed to develop industry, agriculture and harbours was also increasing, leading to considerableland claim in the estuary at this time. Hence the construction of sea walls (Case Fig. 7.1a(iv & v).

(c)

Case Fig. 7.1 (c) Stair-like descent of marsh surfaces towards the present channel. This image shows the vegetated surfaces of (from right to left) the Rumney, Awre and Northwick sediments (marsh edges arrowed) (Littleton Warth, Severn estuary).

ESC07 9/7/06 3:19 PM Page 241

242 PETER FRENCH

Erosion of vegetated surfaces does not justinclude uniform retreat following wave attack.Stripping of the vegetation surface and the expo-sure of underlying sediment to erosion is anotherway in which marsh loss can occur (Fig. 7.10). Itis well accepted that vegetation roots are effec-tive in increasing sediment resistance to erosion.

These roots, however, penetrate only to a certaindepth, for example, up to 1 m for Spartina andup to c. 20 cm for Salicornia, and if the vegeta-tion is largely the same across the marsh surface,then root depth will be uniform. Figure 7.10shows the effect of this at Silverdale marsh in Morecambe Bay, north-west England. Storm

Despite the great thicknesses of sediment deposited, the period over which the RumneyFormation accreted was relatively short. By the end of the nineteenth century, the Rumneymarshes were being cut back by renewed erosion (Case Fig. 7.1a(v)), releasing further quantitiesof sediment into the estuary but, importantly, sediments which were now contaminated withpollutants. This period of reworking, however, lasted just a few decades as by the early twentiethcentury the estuary had switched to being accretional again, with the deposition of the AwreFormation. These sediments again show an increase in coarseness over their predecessors, withsandy clays to clayey sands dominating these deposits (Case Fig. 7.1a(vi)). Importantly, how-ever, the deposition of these sediments incorporated pollution levels present in the water bodyfrom continued industrialization as well as those reworked from the Rumney Formation. Thismarsh unit, however has only minimal development in the estuary because by the 1930s therewas further erosion (Case Fig. 7.1a(vii)) before accretion began again in the 1960s. At this time,the last recognizable marsh unit present in the estuary, the Northwick Formation, began toform (Case Fig. 7.1a(viii)) with the deposition of silty clays and silty sandy clays. At the presenttime, the estuary is eroding again, cutting back into the Northwick deposits.

The causes of this clear erosion and accretion cyclicity are not fully understood, but they docorrespond to periods of increased storminess in the estuary, when larger waves from the pre-vailing westerly wind direction enter the estuary and cause erosion to dominate. Interestingly, theperiod of time between erosion and accretion cycles appears to be shortening, possibly a responseto deepening water caused by a combination of sea-level rise and channel restriction due to landclaim. The appearance of the estuary today is of a series of marshes stepping down towards thechannel, but often terminated in a marsh cliff, several metres in height (Case Fig. 7.1c).

Although this apparent storm-driven erosion and accretion cyclicity is particularly evident inthe Severn, the last decade of the nineteenth century also led to considerable erosion and defencebreaching in many other UK estuaries, such as those of the Essex coast, and the Medway(French 1999). This storm-driven erosion and accretion has important repercussions for humanuse of estuaries and deltas. Once human activity has claimed part of an estuary or delta forsome use, it is considered sacrosanct. Hence, the erosion of these areas as a result of naturalcyclicity inevitably leads to demands for protection and intervention in these natural processes.

Relevant reading

Allen, J.R.L. & Rae, J.E. (1987) Late Flandrian shoreline oscillations in the Severn Estuary: a geomorphologicaland stratigraphical reconnaissance. Philosophical Transactions of the Royal Society of London, Series B 315,185–30.

French, P.W. (1999) Managed retreat: a natural analogue from the Medway estuary, UK. Ocean and CoastalManagement 42, 49–62.

ESC07 9/7/06 3:19 PM Page 242


waves have stripped the vegetation surface fromthe marsh, leading to increased lateral marshretreat (Pringle 1995).

7.3.2 Effects of changes in sediment supply

Although marsh erosion is significant, perhapsthe greatest threat to deltas and estuaries is to reduce the volume of sediment available fordeposition. The relationship between a prograd-ing delta front and an eroding one, or an advan-cing salt marsh as opposed to a retreating one, is closely linked to sediment supply. Figure 7.3shows the complexity of sources and inputs toan estuary or delta region, each varying in import-ance in terms of what they contribute to the sediment budget. Some sources are more easilyvaried than others. For example, hinterlandactivity can change the importance of urban oragricultural runoff, which will lead to increasesor decreases in sediment delivery to the estuary,

whereas port development or coastal defencescan alter the importance of marine inputs.

A reduced sediment supply can also be causedby variations in the current activity which previ-ously delivered that sediment to its site of deposi-tion. For example, in a delta, a main channel maydeliver its load of sediment to one part of thedelta front. Were the bulk of that water supplyto switch, and start to flow to another part of thedelta, the original section of delta front wouldlose its sediment supply and could cease to func-tion as an active part of the delta, potentiallybecoming erosional. The diverted sediment supplycould lead to the formation of new active areasof delta accretion. Over time, therefore, a deltamay reveal a series of accretional phases, eachbeing marked by a delta lobe. A good exampleof this is the Mississippi, a large delta systemwith a series of lobes dating from c. 7500 yr BP(Fig. 7.11). In intertidal parts of estuaries, a similar process of current change as a result of

Fig. 7.10 Erosion of salt marshes caused by the stripping of protective vegetation. Note how the salt marsh vegetation has been peeledback from the sediment surface, exposing the underlying sediments to wave attack (Silverdale salt marsh, Morecambe Bay, UK).

ESC07 9/7/06 3:19 PM Page 243

244 PETER FRENCH

channel switching can occur. At the mouth of theKent estuary in Morecambe Bay (UK), the mainflood-tide channel used to focus sediment accre-tion towards the eastern side of the mouth, wheremarshes developed rapidly and grew seawards atthe rate of several metres a year (Pringle 1995).On the opposite side, little sediment was depositedand marshes were absent. Around the early1970s, this situation changed following theblocking of old channels and the opening of newones, to produce a situation of preferential sedi-ment delivery on the western side of the estuarymouth. This led to the growth of salt marsheshere, and the cutting off of sediment delivery to the eastern side, which is now experiencingerosion in the order of several metres a year(Pringle 1995) (see also Fig. 7.10).

7.3.3 Response of intertidal sediments to sea-level fluctuations

Although loss of sediment input and changes incurrents can be caused by a variety of individualfactors over a range of time-scales, sea-level risehas the potential to alter the entire function-ing of estuaries and deltas because the wholebasis for vegetation establishment, tides andwave energy will change. For example, the chan-nel network in the Rhine–Meuse delta system(Germany) has shown several fundamentalchanges over the past 9000 years as a result ofchanging sea-levels (Törnqvist 1993, 1994; Beets& van der Spek 2000; Berendsen & Stouthamer2000; Woodroffe 2003) and these have beenlinked to a fundamental change in how the

NewOrleans

Baton Rouge

Gulf of Mexico

Prairie Terrace

Sale Cypremort (7500–5000 yr BP)

Teche (5500–3800 yr BP)

St Bernard (4000–2000 yr BP)

Lafourche (3500–1500 yr BP)

Plaquemine (1000–800 yr BP)

Balize (modern)

N

0 20Kilometres

Delta lobes:

Fig. 7.11 Representation of delta lobe formation in the Mississippi delta. (Compiled and modified from E.C.F. Bird (2000) andWoodroffe (2003).)

ESC07 9/7/06 3:19 PM Page 244


system functions. Around 7000 years ago, thechannel pattern changed from meandering toanastomosing. This led to the formation of sedi-ment bars within channels which, as a result,frequently became blocked and changed course(Törnqvist 1993). Consequently, the method ofdelivery of sediment to the delta front becameless reliable. Larger, more permanent meander-ing channels deposited sediment at fixed pointsaround the delta front, but a frequently changingsystem of delivery in an anastomosing channelnetwork meant that the point of sediment deliverybecame less certain and dependable. Around 4000years ago, the system reverted to meandering.

The extent to which sea-level rise will causean impact will depend on the ability of the sys-tem to respond. Bruun (1962) produced a modelby which to predict shoreline behaviour undersea-level rise (see Chapter 8 for a discussion of

its validity). In essence, a shore profile, such as amudflat to salt marsh sequence, will erode andreform so that it remains at the same positionrelative to the tidal frame. Such movement isseen on the ground as an upward and land-ward relocation of vegetation zones. In manysituations, however, this landward movement isrestricted by coastal defence structures, mean-ing that although sea-level rise is forcing the seaward marsh edge landwards, the landwardedge cannot move into the hinterland (Fig. 7.12).This process is known as coastal squeeze.

As salt marsh vegetation communities developat an elevation controlled by the depth, periodand frequency of tidal inundation (Adam 1990),different plant species can tolerate differentperiods and frequencies of inundation. Thosethat can tolerate the greatest inundation occuron the lowest parts of the marsh, whereas those

Fig. 7.12 Restricted salt marsh development along the southern banks of the Humber estuary, UK. The man-made embankment willprevent landward migration of salt marshes as sea-levels rise leading to loss of marsh area. (Photograph by Chris Perry.)

ESC07 9/7/06 3:19 PM Page 245

246 PETER FRENCH

that can tolerate only short or periodic inunda-tion occur on the highest marsh. Figure 7.13ashows an idealized section across a marsh withlow, mid- and high marsh communities. In asimplistic sense, low marsh vegetation is stablebetween the depth (d) and inundation fre-quencies (f) depicted by dfInt, dfLow, where dfInt

represents the upper limit of the intertidal mud-flat and dfLow represents the upper limit of thelow marsh. Similarly, the mid-marsh commu-nities are stable between dfLow and dfMid, wheredfMid represents the upper limit of the mid-marsh.Landward of this point, higher marsh and transi-tional communities dominate. Under conditions

Zone of conversionof terrestrial toaquatic habits

Migration ofhigh water mark

landwards

MHWS'

MHWSS.L.R

MHWS'

MHWSS.L.R

Loss of high marshand upper mid-marsh

Marshedge

Marshedge= lost marsh areas MHWS = Mean High Water of Spring Tides

dfint

High water markfixed at embankment

Terrestrial

Intertidalmudflats

Highmarsh

Midmarsh

Lowmarsh

Transitioncommunities

Intertidalmudflats

Mid-marsh

Lowmarsh

(a)

(b)

Pre-S.L.R

Post-S.L.R

Pre-S.L.REmbankment

Post-S.L.R

Highmarsh

df int

dflow

df lowdfmid

dfmid

dfintdf intdflow

df lowdfmid

Fig. 7.13 Impacts of sea-level rise on estuarine intertidal zonation. (a) No landward constraints allow a gradual shift in intertidal zonesand maintenance of vegetation community succession. (b) The sea wall to landwards results in coastal squeeze and loss of habitats.(Modified from French 2001, p. 278.)

ESC07 9/7/06 3:19 PM Page 246


of sea-level rise (MHWS to MHWS′), waterdepths in the estuary increase meaning that allzones of the marsh are covered more frequentlyand for longer periods (Fig. 7.13a). Clearly thiswill lead to plants becoming unstable in theirnew conditions causing die-back and regrowthfurther inland where water depths and periodsof inundation match those of pre-rise conditions.Hence, the junction between low and mid-marsh communities moves inland to the depth itwas at prior to the onset of sea-level rise. In thecase of the scenario in Fig. 7.13a, this positionmoves from dfInt to df Int, from dfLow to dfLow andfrom dfMid to df Mid. With continued sea-levelrise, this relocation is ongoing and gradual, onlyceasing when sea-level reaches a new stable level.Evidence for such landward migration can beseen in Maryland, USA, where freshwater forestcommunities are being replaced by salt marsh(Darmody & Foss 1979); and in ChesapeakeBay, where terrestrial meadows are being invadedby halophytic vegetation (Bird 1993). In both of these examples, the process is represented inFig. 7.13a by the landward movement of thehigh-marsh–terrestrial communities limit.

When an estuary is in its natural, undefendedstate, any long- to medium-term rise in sea-levelcan be compensated for by these landward‘shifts’ in vegetation zonation provided thatsuitable back-marsh areas for colonization exist(Fig. 7.13a). As soon as sea defences are built,however, any landward shift in these zones isrestricted by a physical barrier (Fig. 7.13b), whicheffectively fixes the high water mark. Therefore,with time, although low and mid-marsh com-munities can shift landwards, there is no spacefor the high marsh, which becomes squeezedagainst the sea wall and will eventually disappear(Fig. 7.13b). As the depth, frequency and periodof inundation increase further, this process willcontinue, with the possibility that a marsh willactually revert to lower marsh species, or evenmudflat. Although such a loss has major implica-tions for the ecology of an estuarine system, itcan also have implications elsewhere in the sys-tem. It has already been seen (section 7.1.2) thatvegetation plays an important role in the trap-ping of sediment and promotes vertical marsh

accretion. With the loss of such vegetated sur-faces, less sediment will be trapped, and so morewill be retained in the water body. This surplussediment has, in some situations, caused addi-tional problems. In Tampa Bay, Florida, forexample, increased turbidity caused by increasedsuspended sediment has reduced plant growthelsewhere in the estuary.

Rising sea-levels are not the only cause ofmarsh loss as vegetation may die back for otherreasons. One way, which remains largely un-explained, is the natural die back in Spartinamarshes. These are low marshes, in that they are the first real vegetation to form as the marshdevelops from the mudflat. In the 1930s, Spartinamarshes along the south coast of England startedto die, leaving large areas of bare mud. In many areas this has been replaced by other lowmarsh vegetation, such as Zostera (Haynes &Coulson 1982; Adam 1990). Although one speciesreplacing another may not be seen as a prob-lem, especially when ecologists favour Zosterabecause of its increased biodiversity and betternutrient dynamics, Spartina is by far the moreefficient species in terms of ability to trap andretain sediment (French 2001).

Although sea-level rise is very much a globalissue, some coastlines are experiencing a rela-tive sea-level fall. Whereas a sea-level rise willreduce the space for estuaries and deltas to occupyunless they can relocate inland, a sea-level fallwill make more space available and, where sedi-ment input remains high, allow expansion of the sedimentary and vegetation zones. The deltafront will move further onto the continentalshelf, and intertidal zones in estuaries will startto encroach on the original intertidal area andshift seawards.


ACTIVITIES IN DELTAS AND ESTUARIES

Although estuaries and deltas are subject to con-siderable natural sediment dynamics, they alsorepresent some of the most intensively exploitedenvironments. The main reason for this is theirproximity to, yet shelter from, the open sea, and

ESC07 9/7/06 3:19 PM Page 247

248 PETER FRENCH

their suitability for industrial and agriculturaldevelopment. Human activity has, therefore, pro-duced significant modification to many systems(Table 7.2).

7.4.1 Anthropogenic impacts on rates and stylesof sedimentation

Reduced freshwater flow will produce signific-ant changes in the functioning of deltas andestuaries. In the former, reduction in sedimentsupply can lead to increased marine influence,and thus increased erosion, whereas in the latter,an estuary may experience changes in flowstructure, and ebb/flood dominance. One of themain ways of affecting the amount of riverineflow is by the construction of dams. Completionof the Aswan High Dam in 1964, for example,reduced the supply of sediment to the Nile deltafrom 124 to 50 million tonnes a year (Carter1988). The net result was a dramatic increase in coastal erosion along parts of the delta edge(see Case Study 7.2). In a similar way, closure of the Akosombo Dam in 1961 completelyblocked sediment supply to the Volta delta inGhana (Ly 1980). Apart from initiating largeamounts of coastal recession, the salt wedges inthe delta channels moved further landwards,altering local ecology. Furthermore, pollutantflushing was reduced leading to increased resi-

dence times for pollutants in estuarine channels,and sediment deposition patterns were modifiedby weakened ebb currents (Collins & Evans1986). The only sediment that now reaches thisdelta comes from erosion of sediments downstream of the dam. Dams on the Ebro (Spain)and the Rhône (France) have also reduced sedi-ment supply to coastal deltas by 96% and 90%respectively (Viles & Spencer 1995).

As well as affecting the amount of sedimententering the estuary or delta, dams also retainfresh water, thus reducing the quantity thatenters the estuarine area. Stratified and partlystratified estuaries (see Fig. 7.1) rely on fresh water to drive their circulation and create thestratification. This freshwater-induced circula-tion also controls the movement of nutrients onwhich primary and secondary productivity ofthe system depends (Mann 2000). More funda-mentally, many brackish water species have,through long-term adaptation, adjusted to copewith seasonal fluctuations of freshwater input,caused by seasonality of flow from river basins.One impact of dams is that they often reduce oreliminate this seasonality, in that water is storedat times of high flow, and released at times oflow flow.

Whereas dams are built to constrain fresh waterin rivers, barrages are built across estuaries towithhold water on the ebb-tide for the purposes

Table 7.2 Human activities and related impacts in estuary and delta environments.

Activity Impact

Land claim Loss of intertidal habitat, coastal squeeze, increased need for flood defencesCoastal defence Loss of intertidal habitat, increased erosion, changes to natural sediment cycling,

coastal squeeze, alteration to wave and tidal processesTourist development Increased visitor pressure, habitat loss, increased contamination/pollution,

increased need for coastal defencesIndustry Contamination/pollution, habitat loss, increased need for defences, increased need

for shipping access (dredging)Barrages Loss of intertidal habitat, changes to tidal regime, contaminant retention, greater

brackish water penetration up river, sediment retention, changes to water tableSediment extraction Loss of submarine habitat, changes to tidal and wave currents, loss of wave

protection, re-suspension of sediment and remobilization of pollutantsWaste disposal Contamination/pollution, increased turbidityDams Reduction in freshwater supply/increase in net salinity, reduction of sediment supply

from catchmentCatchment land use changes Changes in sediment supply, changes in pollution levels, changes in freshwater input

ESC07 9/7/06 3:19 PM Page 248


Case study 7.2 The impact of dams on deltas and estuaries: the Nile delta and the Aswan High Dam

BACKGROUND

The Aswan Dam was built in order to provide irrigation and drinking water, via the constructionof a large reservoir (Lake Nassar), to the Nile catchment, and also to facilitate the developmentof hydroelectric power to the growing population of Egypt (Rashad & Ismail 2000). Such asupply was considered vital for a country whose population (c. 48 million) had complete depend-ence on the Nile. Although its completion in 1964 facilitated this, environmental impacts werealso common, not least in the delta area downstream of the dam. This was not the first structureto be used to regulate water along the Nile, however, with various schemes, such as constructionof the Aswan Low Dam in 1902, having gradual accumulatory impacts on the delta.

As with any coastal landform, deltas undergo alternate constructive and destructive episodes(see Fig. 7.11 for the example of the Mississippi). Cores taken in the area have shown that therehave been a series of deltas at the mouth of the Nile, which have accreted and eroded throughoutPliocene and Pleistocene times (i.e. over the past 5 million years) (Stanley & Warne 1998), withthe present delta starting to form c. 8000 to 6500 years ago. In the early stages of the formationof the present delta, seven major channels drained across the delta surface, each with its ownlobe at the coast (Badr & Lofty 1999). Significantly, by 2000 years ago, two of these had dis-appeared, and by 1000 years ago, a further three had gone (McManus 2002), leaving the twowe now recognize as the Rosetta and Damietta. This loss of channels was partly the result ofnatural evolution of the delta, but also early attempts at irrigation and management.

Over the past 150 years, the delta has been in an erosional phase (Stanley & Warne 1998)with the majority of the 260 km long delta face undergoing erosion and landward retreat (Frihy1996). Only small areas of localized accretion now occur, such as in association with jetty construction at Damietta Harbour (El-Asmar & White 2002) (Case Fig. 7.2). The reason forthe widespread erosion has, to a great extent, been more organized water regulation, which has affected the balance between sediment input, coastal sediment drift and subsidence in thedelta area. Despite earlier management efforts, the Aswan High Dam has, arguably, had thegreatest impact on delta decline. Pre-closure in 1964, the Nile supplied c. 124 million tonnes ofsediment a year to the delta area (Carter 1988). Such a flow of both water and sediment wassufficient to maintain the two major channels, but also a series of minor distributary channels,which provided an annual deposit of fertile silt on farmland. Following closure, the amount ofsediment delivered per year dropped to c. 50 million tonnes (Carter 1988). This sediment wasnotably finer grained than pre-closure (Stanley & Wingerath 1996), and largely derived fromthe river downstream of the dam, and also from the delta hinterland itself. More significantly,however, the volume proved insufficient to compensate for that lost to erosion. As a result, thedelta front became net erosional.

As well as reduced sediment load, the reduced river flow led to the silting up of many of thedistributary channels, leading to major changes in delta morphology and ecology (Case Fig. 7.2).For example, reduced discharge and the easterly drift of sediment at the coast blocked many of the smaller estuaries (Frihy 1996). Chesworth (1990) shows that of the c. 55.5 billion m3 ofwater released from the dam, only 17.5 billion m3 reaches the delta coast, the rest being eitherextracted or lost to evaporation. Of the two main channels, the Damietta and Rosetta, the former is generally dry for much of the year, and the latter carries the bulk of what flow remains(Bird 1985).

ESC07 9/7/06 3:19 PM Page 249

250 PETER FRENCH

F

B

S

Branc

h

LSD predominantly easterly

1 A3

A

4

A2

B

Water QualityS = SaltB = BrackishF = Fresh

See table for descriptons for letters, numbersand roman numerals

I

II III

Lake Idku

Lake Burullos

Ro

setta

Bra

nch

Dam

i ett

a

LakeManzalah

Su

ez

Ca

na

l

Cairo

NileR

iver

0 50km

N

Sand barrier

Former distr ibutaries

Marshland

M e d i t e r r a n e a nS e a

Erosional sectorsSite Erosion rate (m yr−1) Measurement period Reference1 >50 1909 to late twentieth century Stanley & Warne (1998)

>100 1971 to mid-1990s Frihy (1996)113 1985–1991 White & El Asmar (1999)52–88* 1971–1990 Frihy et al. (2002)3–13† 1990–2000 Frihy et al. (2002)118.6 1984–2000 Ahmed (2002)18–230 Last century Ahmed (2002)

2 Unspecified 1909 to late twentieth century Stanley & Warne (1998)10 1909 to mid-1990s Frihy (1996)10 1971–1990 Frihy et al. (2002)15 1984–1991 White & El Asmar (1999)

3 6.5 1971 to mid-1990s Frihy (1996)1.3 1909–1989 El-Fishawi (1994)2.1 1984–1991 White & El Asmar (1999)

4 6.5 1971 to mid-1990s Frihy (1996)3.8 1909–1989 El-Fishawi (1994)5.0 1984–1991 White & El Asmar (1999)

*Pre-defence construction in 1990. First figure to west of promontory, second to east.†Post-defence construction. First figure to west of promontory, second to east.

Accretional sectorsSite Accretion rate (m yr−1) Measurement period ReferenceA Up to 13 1971 to mid-1990s Frihy (1996)

14 1971–1990 Frihy et al. (2002)B Unspecified 1909 to late twentieth century Stanley & Warne (1998)

Offshore sediment transportSite To offshore (m3 yr−1) Measurement period ReferenceI 3.2 × 106 Current McManus (2002)II 1.48 × 106 Current McManus (2002)III 1.8 × 106 current McManus (2002)

Case Fig. 7.2 The Nile delta showing areas of erosion and accretion: LSD, longshore drift. (Modified from Carter 1988, p. 483.)

ESC07 9/7/06 3:19 PM Page 250


of power generation, or for amenity uses. Theirimpact, however, is just as severe. By reducingthe mixing of fresh and salt water, barrages canaffect the salinity structure of the estuary, suchas occurred at the Rance tidal power station in France (Shaw 1995). Sediment retention mayalso increase upstream of the barrage, as itwould in a dam reservoir, such that the sediment

budget may be affected downstream. The exactnature of this impact will depend on whether the estuary is ebb or flood dominant and thuswhether the prime sediment source was land-based or marine (see section 7.2.2). Changes in water movement and reduction of fresh-water flow may also cause contaminant levels tochange in response to water storage upstream of

Over the past 150 years, the Nile delta has thus changed from a constructive, wave-dominateddelta that had formed over the past 7000 years, to an erosive coastal plain (Stanley & Warne1998). This has been chiefly caused by a negative sediment budget that is insufficient to main-tain the relative height of the delta surface following subsidence. With increasing developmentand population growth, including the construction of a new road along the seaward margins ofthe delta to open up less accessible areas for development (Frihy 1996), delta degradation islikely to continue, with increased subsidence and greater risk of marine inundation and flooding.Furthermore, contamination of groundwater is likely to increase and to move further inland.The future is, therefore, bleak. The only way that the Nile can regain its role as a functioningdelta is to reinstate water flow and sediment supply. In reality, this is not likely and will result incontinued shoreline erosion and land subsidence.

Relevant reading

Ahmed, M.H. (2002) Multi-temporal conflict of the Nile delta coastal changes, Egypt. Proceedings of theConference Littoral 2002: the Changing Coast, Porto. Eurocoast/EUCC Portugal 2002.

Badr, A.A. & Lofty, M.F. (1999) Tracing beach sand movement using flourescent quartz along the Nile deltapromontories, Egypt. Journal of Coastal Research 15, 261–5.

Bird, E.C.F. (1985) Coastline Changes: a Global Review. Wiley, Chichester.Carter, R.W.G. (1988) Coastal Environments. Academic Press, London.Chesworth, P.M. (1990) The history of water use in Sudan and Egypt. In: The Nile (Eds P.P. Howell &

J.A. Allan), pp. 65–79. Cambridge University Press, Cambridge.El-Asmar, H.M. & White, K. (2002) Changes in coastal sediment transport processes due to construction of

New Damietta Harbour, Nile Delta, Egypt. Coastal Engineering 46, 127–38.El-Fishawi, N.M. (1994) Relative changes in sea level from tide gauge records at Burullus, central part of the

Nile Delta coast. INQUA MBSS Newsletter 16, 53– 61.Fahim, H.M. (1981) Dams, People and Development: the Aswan High Dam Case. Pergammon, Oxford.Frihy, O.E. (1996) Some proposals for coastal management of the Nile delta coast. Ocean and Coastal

Management 30(1), 43–59.Frihy, O.E., Debes, A.D. & El-Sayed, W.R. (2002) Processes reshaping the Nile delta promontories of Egypt:

pre- and post-protection. Geomorphology 53, 263–79.McManus, J. (2002) Deltaic responses to changes in river regimes. Marine Chemistry 79, 155–70.Rashad, S.M. & Ismail, M.A. (2000) Environmental impact assessment of hydro-power in Egypt. Applied

Energy 65, 285–302.Stanley, D.J. & Warne, A.G. (1998) Nile delta in its destructive phase. Journal of Coastal Research 14(3),

794–825.Stanley, D.J. & Wingerath, J.G. (1996) Nile sediment dispersal altered by the Aswan High Dam: the kaolinite

trace. Marine Geology 133, 1–9.White, K. & El-Asmar, H.M. (1999) Monitoring changing position of coastlines using thematic mapper

imagery, an example from the Nile Delta. Geomorphology 29, 93–105.

ESC07 9/7/06 3:19 PM Page 251

252 PETER FRENCH

the barrage. On the one hand, riverine inputsmay achieve a greater dilution owing to theincreased volumes of water stored in the tidalreservoir, but on the other, changes in waterdepth, current activity, flushing and oxygen levels can all affect contaminant breakdown and flushing. In the Tawe estuary, South Wales,flushing has been reduced significantly sincebarrage construction. Although incoming watermay overtop the barrier, being denser than freshwater, a captive salt wedge has formed insidethe barrage, which has led to stagnation behindthe barrage during neap-tide periods (Dyrynda1996). Also in the Tawe, contaminants havecontinued to enter upstream of the barrage,leading to organic enrichment, algal blooms andmetal enrichment. Overall, the water quality has deteriorated significantly and has led to theneed for artificial remediation, such as periodicdraining of the lake through sluices.

Other issues relating to tidal barrages includemodification to currents and local increases insea-level. Estuarine ecosystems owe their exist-ence and spatial variation to the frequency andperiod of tidal inundation. Barrages will changeall of these parameters and, hence, will affect the flora and fauna of the system. In many situations there will be an increased segregationbetween marine and riverine communities. Inthe case of La Rance in France, the constructionphase (1963–66) saw the complete destruction ofthe estuarine ecosystem due to the sealing of theestuary. However, post-commissioning restock-ing has allowed a new, if not ‘natural’, ecosystemto develop (Rodier 1992).

Although dams and barrages provide thegreatest impact on the overall functioning ofdeltas and estuaries, other human activities canalso lead to major changes. Tides, waves, riverflow and currents all combine in the function-ing of an estuary or delta, and thus any activitythat changes these has the potential to producesome form of modification to process and form.Such activities include shore-normal structures(ports, jetties), bridges, coastal defence struc-tures, wrecks, dredging, land claim, wetlandcreation (managed realignment) and changes inland use.

7.4.2 Reworking of contaminants

Section 7.2.3 and Fig. 7.8 demonstrate howcontaminants can become attached to sedimentgrains, and thus become incorporated into thesediment record (see e.g. Fig. 7.9). One key issuefor water quality is how stable these contam-inants remain, and whether they may re-enterthe water body from the host sediment at a later date. Oil, for example, will break down in the aqueous environment. When treated withdetergents to break up a slick, however, the oilresidues often sink through the water column to the sea floor and become incorporated intosediments via pore spaces, burrows or throughburial. Similarly, ships that sink will generallytake their contents with them, again transferringoil to the sea floor. During the Gulf War of the early 1990s, around 14 million barrels of oil were released into the Gulf environment(Alam 1993). Some of this oil, particularly thatfrom wrecks, was released at the sea floor, andmuch of this filtered into sediments throughpore spaces and burrows, effectively putting itinto storage. Over the following years, naturalreworking of these sediments caused some ofthis oil to leak out, recontaminating some areas.In section 7.2.3 it was reported that oiled saltmarshes can regenerate after an oil spill, and anexample was cited from Buzzard’s Bay, Florida(Teal et al. 1992). Although this marsh had fullyrecovered after 20 years, a major concern relatesto the fact that large quantities of degraded oilremain within marsh and mud-flat sediments.Even though at the current time this is not aproblem, any process that disturbs these sedi-ments, such as erosion, could well cause therelease of this oil store, estimated to be largeenough to damage fauna (see also Chapter 9).

The storage and future release from sedimentstorage can occur for any of the contaminantsmentioned in this chapter. Contaminants willstabilize to the environmental conditions pre-sent at the time of deposition, and their chemicalform may adjust accordingly through reactionwith other substances, or with organic/mineralsediment grains. If conditions change, however,these contaminants may become unstable, and

ESC07 9/7/06 3:19 PM Page 252


re-enter the system. Such occasions could include:erosion, where material is physically removedfrom a store back to the water body; sea-levelrise, where a rise in water table could destabilizeoxic sediments or increased salinity up river maydestabilize freshwater sediments; and vegetationdeath, where pollutants locked up in vegetationre-enter the system as material decays. Some ofthese processes may be gradual, such as sea-levelrise, but others, notably vegetation die-back or a storm causing a large retreat of a marsh, can produce contaminant flushes, whereby largequantities of stored pollutants enter the waterbody at the same time (see Case Study 7.3).

Such changes in the stability of marshes andtheir constituent pollutants can also be causedby human activities. Increasingly, coastal man-agement is turning to soft engineering to protectthe coastline from flooding. One such techniqueused in estuaries is managed realignment, a pro-cess where previously claimed land is allowed to flood and become intertidal again (see French2001). The implications here are that the sedi-ments forming the claimed land were depositedunder estuarine conditions, and contain levels of pollutants reflecting the estuary at the time ofdeposition. When these sediments were initiallyclaimed, they were drained and became freshwater

Case study 7.3 The recycling of contaminants from eroding salt marshes

Sediments laid down in estuaries and deltas incorporate a range of contaminants which reflectthose present in the water body at the time of deposition (see section 7.2.3). In time, these sediments and contaminants accumulate to produce an environmental record of contaminantstatus of the system (see Fig. 7.9 and Case Fig. 7.3). However, when these sediments start toerode, the contaminants contained within them will be reintroduced to the water body. Thiscase study represents an assessment of the contaminant input from one such eroding marsh.

The Severn estuary (UK) is a large, macrotidal estuary that has undergone a series of erosionand accretion phases (see Case Study 7.1). At the current time, these sediments are eroding, re-introducing sediments and contaminants to the estuary. The marsh (Case Fig. 7.3) represents a1.12 m sequence of sediments of the Northwick Formation, located on the southern shore of theestuary at Northwick Warth. The marsh was sampled via 1-cm slices, each of which was analysedto determine levels of copper, lead and zinc (full analytical method is given in French 1996).

The resulting data set provides a record of the levels of contaminants present in each succes-sive 1-cm slice, and thus it is possible to estimate the quantities of these materials that will re-enter the estuary as this marsh retreats landwards. It is known that the average retreat rate of this marsh is 0.17 m yr−1, and that the length of the eroding frontage is 3.2 km. It has to beassumed that the marsh retreats uniformly landwards, and that the height is constant through-out the length. These assumptions having been made, the volume of eroded sediment (in cubicmetres) from each 1-cm (0.01 m) slice is 5.44 m3 yr−1. To determine the levels of metal input, itis necessary to consider the concentrations within each of the 112 × 1-cm-thick slices. For thetopmost layer (0–1 cm), analysis has shown that there is 43 ppm Cu, 96 ppm Pb and 300 ppmZn (French 1996). Using these values, the volumes of sediment eroded from this topmost sliceare adding 248 g of copper, 554 g of lead and 1730 g of zinc per year.

Using this approach to calculate the amount of copper, lead and zinc in each layer, thenadding the total for each together, the salt marsh at Northwick Warth is contributing 33,953 gof copper, 80,707 g of lead and 239,294 g of zinc, per year, back to the waters of the Severnestuary. This analysis will not show in what form these metals occur, in terms of chemical stability or bio-availability, but it does present an amount that can be used in determining theEnvironmental Quality Standard (see section 7.5.2). The key factor, however, is that the salt

ESC07 9/7/06 3:19 PM Page 253

254 PETER FRENCH

dominated. This would have represented the firstmajor stability change. Managed realignmentreintroduces brackish water to these areas,increasing the saturation of the soils and causingfurther remobilization of many pollutants. Theprocesses by which remobilization occurs arediverse and include metal transformation by pro-cesses such as methylation, change of chemicalphase due to the changing oxygen availability,salinity, redox, bioturbation and acidity/alkalinity.Bryan & Langstone (1992) provide a full review

of these processes, and Emmerson et al. (2000)demonstrate the role of changing acidity andchlorinity in the release of sediment-bound metalsin a field-based situation at Orplands, Essex, UK.Blackwell et al. (2004) look at the immediatepost-breaching change in the Torridge estuary,Devon, UK.

Problems with mobilization can become severeif levels of a particular contaminant are pres-ent in sediments in high concentrations. Thismay be the case where sediments are associated

marsh investigated here is only a small part of the salt marsh resource of the estuary. The twobanks of the Severn estuary combined cover a distance of over 400 km, much of which is saltmarsh. Hence, the release of material from these sources is potentially large.

Relevant reading

French, P.W. (1996) Long-term variability of copper, lead and zinc in salt marsh sediments of the SevernEstuary. Mangroves and Saltmarshes 1(1), 5968.

Case Fig. 7.3 Eroding sediment of the Northwick Formation, Northwick Warth, Severn estuary. The erosion of this marsh,accreting since the 1930s, is now contributing stored contaminants back in to the water body. (Data from French 1996.)

ESC07 9/7/06 3:19 PM Page 254


with industrial activity. Fraser (1993) reports ondredging operations in New Bedford Harbour(USA), where sediments had been contaminatedwith waste from the adjacent electronics indus-try. Particular concerns focused on PCBs andmetals, PCB levels being the highest recorded in any estuary in the USA. When the sedimentsremained in situ, these contaminants remainedstable and locked up in the sediments. As such,they were not considered an environmental threat.Natural reworking, however, led to the reintro-duction of some material and, as a result, fishingand shell fish collection were stopped. Dredging,however, threatened large-scale destabilizationof the stored contaminants and the large-scalereintroduction to the water body. As a result,dredged material was treated as high grade in-dustrial waste and had to be containerized andisolated from the environment. Such situationsare rare, but the New Bedford Harbour examplehighlights some interesting issues relating toestuarine and delta sediments. Most notable are the concerns of reintroduction and reactiva-tion in the contemporary environment (see CaseStudy 7.3). Section 7.5 looks at management andthe idea of emission quotas. Although this workswell for industrial point source emissions, it oftenfails to adequately account for the release ofcontaminants through sediment reworking.


Humans have continually strived to alter estuar-ies and deltas to facilitate their own needs. Eversince industrialization began, intertidal sedimentshave been receiving increased quantities of contaminants, and agricultural, industrial andurban growth have been claiming large areas of the intertidal zone, and development in theseregions has necessitated greater amounts ofcoastal defence and flood protection. The Teesestuary is a classic example of an area that hasundergone progressive land claim for industrialdevelopment (Fig. 7.14). In the mid-nineteenthcentury, the estuary contained large areas of salt marsh and intertidal flats yet by the mid-1970s, c. 3300 ha, or 83% of this area had been

claimed for industrial and port activity (Davidsonet al. 1991). Through this process, the intertidalareas of many estuaries and deltas have becomehighly altered in terms of morphology, function-ing and sediment/water quality.

The range of impacts associated with this act-ivity is large and diverse, and poses potentiallyserious problems for managers (see Table 7.2).In addition, many of these impacts are accumu-latory. For example, land claim may be carriedout in a piecemeal way over several centuries,but overall represents a large collective loss ofsalt marsh and mud flat. Although pressures forindustrial and port development may be increas-ing, other threats are diminishing. The change inagricultural economics in the 1980s and 1990shas reduced the threats posed by agricultural landclaim and, indeed, this legacy is actually seeing a reversal as agricultural land is returned to theintertidal zone through managed realignmentschemes. In addition, increased environmentalprotection measures aimed at safeguarding inter-tidal areas for their wildlife interest have grownin stature, and, significantly, people have increasedunderstanding of how these systems work andfunction. This means that it is now possible to gain a greater understanding of how systemschange over time in relation to external forces,such as sea-level rise, and how they may react tohuman-induced changes, such as dam construc-tion and land clearance. Once understanding ofaction–impact relationships develops, manage-ment becomes easier as it can start to be proactive,rather than reactive.

These human-induced changes mean thatestuaries and deltas are now afforded a higherdegree of protection than has hitherto been thecase. Although this may be fine for future man-agement, it does not help overcome some of thelegacies left to modern managers by previousland-use policies. In considering this historicallegacy, it becomes necessary to think of manage-ment in a variety of ways. Previous changes todeltas and estuaries, such as dam constructionor the claiming of salt marshes, have resulted in considerable adjustments to the natural sys-tems. Therefore, it is critical at an early stage in the management cycle to decide whether

ESC07 9/7/06 3:19 PM Page 255

256 PETER FRENCH

management should be a means of maintaininga form of status quo (i.e. protect what these systems have become and prevent further human-induced change), or whether it should be a meansof managing what they once were, i.e. to regaina natural estuary or delta. The latter is imprac-tical in many situations owing to the extent towhich humans have developed and interfered withnatural functioning. The former is perhaps themost workable, but the issue then becomes whatto take as the status quo. This debate becomescomplicated. For example, the Nile delta is still

changing in response to the construction of theAswan High Dam, so do managers wait until it has stabilized, or begin to manage the area asit is at the present time, given that the currentsituation is only a snap-shot representing oneshort section of the conveyor belt of gradualadjustment and change?

The key debate here is the idea of conservingor preserving. Natural systems need to changeand adapt to changing environmental condi-tions. Whether such changes are natural oranthropogenic is, to some extent, unimportant.

(a) (b)

(c)

Middlesborough

Seal Sands BranSands

Coatham Sands

GreathamCreek

CowpenMarsh

1km

Intertidal land

Redcar

1853

18811972

SealSands

1974

1881

1921

1972

19381740–1808

1906 1906

1921

1953

1953

1972

1972

1881

1938

1972

BranSands

North GareSands

1953

1953

1938

1906

1723

3000

2500

2000

1500

1000

500

01850 1900 1950 1990

Year

Inte

rtid

al a

rea

rem

aini

ng (

ha)

Tees Estuary: land claim

Fig. 7.14 The sequential claiming of the Tees estuary, UK: (a) Tees estuary in the mid-nineteenth century; (b) phases ofsubsequent land claim; (c) loss of intertidal area between 1850and 1900. (Modified from Davidson et al. 1991.)

ESC07 9/7/06 3:19 PM Page 256


The key thing for management today is to pro-tect environments from further avoidable change(i.e. caused by human actions), and to allowthem to adjust to ongoing forces of change, i.e.sea-level rise. Hence, management is about stop-ping issues such as land loss and contaminantinput, and providing for other needs, such asincreased sediment to compensate for negativebudgets and the freeing up of land to allow it torespond to sea-level rise. To do this, it is neces-sary to determine what the anthropogenic forcesof change are. Viles & Spencer (1995) present a case study relating to wetland loss on theMississippi delta, in which they cite a range ofmanagement issues. These are split into threecategories: ‘origin and evolutionary development’;‘degradation and loss’; and ‘human activities’.The former includes issues such as storm activ-ity and channel switching, which it could beargued are natural processes that form part ofthe ongoing evolution of the delta. The secondcontains management issues that arise from the first, such as loss of sediment due to channelswitching. The final category contains the impactsof human activities, such as dam building, landclaim and drainage. In terms of managementapproaches, the final and to some extent the second categories are management problems,but the first is not. It is common, however, andnot just in the Mississippi, that the first categorybecomes a management issue because these processes affect human development. In otherwords, it is necessary for processes of naturalsystem evolution to be managed purely becauseif such evolution were permitted, it could havean impact on human activity.

This then puts the earlier point of whether to conserve or preserve into another context.Frequently the overriding strategy for manage-ment is that although it is considered a goodthing, it is only good if it does not inconveniencehuman activity. For estuaries and deltas, and toa great extent coastlines, this is a critical issuebecause often management and remediationcannot progress according to the best environ-mental reasons because of the human interest in an area. It is relatively easy to manage somehuman activities based on our knowledge of

their impacts. For contaminants and pollutants(see section 7.5.1), abatement and input reduc-tion can be achieved easily providing that we areable to identify what these inputs are, and wherethey originate from (see previous discussion oncontaminant reworking, and Case Study 7.3).Other aspects of human interference are moredifficult. The reversal of land claim in order tofacilitate sea-level rise is a case in point becausefreeing up land, often in areas where land use isintense, cannot be facilitated easily. Hence, inmany cases, management will fail purely becausemanagers cannot undo the legacy of the past andit may ultimately be that this legacy facilitatesthe demise of many estuarine and delta envir-onments. One important criterion, however, isthat whatever management we use, it should beaimed at facilitating a natural system, and not at managing natural processes that will furtherdisrupt the system.

7.5.1 Managing for the prevention ofenvironmental change

One of the main difficulties with managementrelates to where the management responsibility forsuch actions lies. One argument is that it shouldlie with interest groups and various government-based environmental authorities. With such anapproach, however, there is significant potentialfor conflicting approaches to management. Forexample, an electricity generating company maymanage a dam and its lake, but the water whichemerges from that lake becomes the responsibil-ity of a river authority. Hence, is the release ofwater based on the demand for electricity gener-ation, or is it on the basis of what the estuary ordelta downstream needs to support extraction,fish populations, or sufficient quantities of waterto dilute permitted waste discharges? Such issuesdemand an overall management strategy cover-ing rivers, estuaries and coasts. However, suchsystems are generally thought of as unwieldy andunworkable. Hence, smaller units of manage-ment have to be used.

The Estuary Management Plan (EMP) canprovide for development and general land-useplanning in the estuary (or delta); it cannot

ESC07 9/7/06 3:19 PM Page 257

258 PETER FRENCH

manage what goes on beyond the estuary limits.For deltas, this is critical because of the issue ofsediment supply: issues that are linked to thefreshwater system, and come under the Catch-ment Management Plan (CMP). For estuaries,what goes on to seaward is also important andagain these fall within another managementplan region, the Shoreline Management Plan(SMP). The risk of having three plans is that the lack of communication between differentmanagement groups could lead to one plan recommending a course of action that could beof detriment to the other. An obvious examplehere would be freshwater management and dam construction as cited above. Managers ofthese respective plans, however, are encouragedto meet and consult on the outcomes of theirrecommended activities, which is generally donethrough a liaison committee that facilitates dia-logue at the planning stage, and also overseesthe running and implementation of the plans(Fig. 7.15).

The formulation of management plans shouldbe a dynamic process that develops and imple-ments a co-ordinated strategy to achieve theconservation and sustainable multiple use ofestuaries and deltas. Inherent in this is the ideaof co-ordination, which should bring catchment,estuarine and shoreline management together.Such co-ordination is facilitated by a liaisoncommittee with the role of promoting dialoguebetween different managers. Eventually, thisshould then fall within the realm of coastal management, of which estuaries and deltas are integral parts.

7.5.2 Managing for the prevention ofanthropogenic inputs

The natural ability of estuaries to assimilatecontaminants means that they can be used totreat some of the waste material generated bysociety. It is important to ensure, however, that the permitted inputs do not exceed the

Liaison Committee

Contaminant inputsLand-use changeFisheriesAmenity useFlood defencesConservationDevelopment

Catchment

Contaminant inputsLand-use changeFisheriesAmenity useFlood & coastal defencesSediment budgetDredgingConservationDevelopment

Estuary & delta

Contaminant inputsLand-use changeFisheriesAmenity usageIndustrial usageSediment budgetCoastal defencesConservation

Shoreline

Will

impa

ct o

n

Will

impa

ct o

n

CatchmentManagement

Plan

EstuaryManagement

Plan

ShorelineManagement

Plan

Sho

uld

impa

ct o

n

Sho

uld

impa

ct o

n

Liaison Committee

Fig. 7.15 Managementstructure and issueshaving an impact uponestuaries and deltas.

ESC07 9/7/06 3:19 PM Page 258


estuary’s capacity to assimilate them. Hence,one method of ensuring safe inputs is to basethis on a determination of how much of a con-taminant a particular estuary can assimilatebefore functioning is impaired. Such a deter-mination is a function of size, tidal range, fresh-water input, floral/faunal/sediment assimilationand flushing ability.

McClusky (1989) makes the general point that‘. . . the solution to pollution is dilution’, wherebycontamination in a water body can have lessimpact, and avoid becoming a pollutant, if it isdiluted by a large enough volume of water. Thus,large estuaries with large tidal ranges and wherelarge volumes of water are exchanged on eachtidal cycle will facilitate greater contaminantassimilation. However, other factors also play animportant part in the calculation of safe limits.The nature of the contaminant, its environ-mental residence time (is it conservative or doesit break down), and the ecology of an area areall important. Hence, an estuary’s ability toassimilate contamination is partly a function ofits size and flushing ability, but importantly it is also a function of its other component parts,such as ecology, the contaminant itself, and alsothe spatial position along the estuary axis atwhich the discharge is made.

To determine how much of a particular contaminant an estuary can assimilate withoutimpairing its functioning, and therefore thedesired state of the estuary post-discharge, isknown as the Environmental Quality Objective(EQO). Having determined this, it is necessaryto determine the levels of discharge which canbe safely discharged without threatening thisobjective. This is known as the EnvironmentalQuality Standard (EQS). Examples would includethe amount of a metal contaminant that couldbe introduced before toxic effects were felt byfauna and flora; how much sewage could beintroduced before the biological oxygen demandincreased to the detriment of fish; or how muchsediment could be introduced before the deposi-tion rate threatened to bury infauna. Havingdetermined safe limits, it is then necessary toproportion this permitted amount amongst thosegroups that wish to discharge.

The EQO/EQS approach is increasingly usedto determine safe levels of inputs into an estuary.There are key issues, however, which althoughappearing simplistic in the procedures of deter-mining the EQO/EQS, are actually problematic.Determining safe levels of discharge to satisfythe EQS is difficult because it relies on measure-ment of the range of estuary-specific factors outlined above, such as natural dispersal andflushing. Here the problem lies with determiningwhat is typical for the system concerned. Atwhat point, for example, do you take measure-ments of tidal range, and what is assumed as anormal freshwater input? The use of mean tidalrange and mean river flow is an obvious answer,but a further problem is that a mean is a com-bination of high and low values, which occur at different times of the year. Tidal range, forexample, varies on a fortnightly cycle with neaps(lower) and spring (higher) tides, and also varieson a seasonal scale (equinox tides). Furthermore,tides also follow longer periodicity cycles. Interms of determining safe discharge limits, usinga mean value will mean that there are times whendischarge could be higher (i.e. when tidal rangeis above the mean), but others when it should belower. The key issue here is that in the case of thelatter, when discharge becomes more concen-trated, will this still be ‘safe’ in terms of estuaryfunctioning? Another key factor, and one that isregularly overlooked, is the amount of materialthat enters an estuary through routes which arenot accounted for in the determination of theEQS. One mechanism is through the reworkingof stored contaminants (see Case Study 7.3). Thiscould, in certain situations, lead to significantinputs above those considered safe both by themanagement authority and by the EQO.

7.6 FUTURE ISSUES

The introduction to this chapter highlighted the fact that estuaries and deltas are formed by different processes. However, once channelsform on a delta surface, processes tend to con-verge, leading to many of the issues pertainingto sediments, human impacts and pollutants being

ESC07 9/7/06 3:19 PM Page 259

260 PETER FRENCH

largely common. Susceptibility to changes in theseissues makes both systems highly vulnerable,both in the long and short term. The last sectiondiscussed management and the prevention ofsome elements of change. Remaining problemsinclude: contaminant management and the legacyof stored contaminants in many historicallyindustrialized estuaries; loss of habitat, largelynow the result of erosion and land claim fordevelopment; freshwater management due to thegrowing conflict between the need for increasedwater resources for a growing population, and theneed to maintain freshwater flow into brackishsystems; and sea-level rise.

Some elements of change, such as erosionaland depositional cycles in marshes and in deltachannels, may show a cyclicity (see e.g. CaseStudy 7.1). There is an increasing element ofdirectional change, however, which could seethese systems under increasing threat in thefuture. The major decision for managers is torecognize the point at which management has to give up trying to maintain a status quo, andallow change to happen. It is a great mistake forcontemporary scientists and morphologists toregard all of our current environments as beingat the end of their evolutionary cycle. It is muchmore likely that much of the current change ispart of a process of evolution and must, there-fore, be allowed to continue for the sake of thefuture stability of the environment. A key factorto determine, however, is to what extent willhuman activity accelerate or alter the rate anddirection of change. This final section will dis-cuss some of these key issues.

7.6.1 Impacts of climate change and sea-level rise

There is some irony in the fact that estuaries represent the drowned mouths of river valleys,as this immediately ties their origin into sea-levelrise, yet today there is concern about the threatof this very same process as an agent of change.Sea-level rise, however, represents one of thegreatest issues in relation to contemporary coastalsystems. Any area that experiences tides willexperience changes caused by deepening seawater and increased tidal penetration inland (see

Chapter 1). This is exacerbated for estuaries and delta channels, which are intensively usedfor development and, because of historic trendsof sea-level rise and of coastal erosion and flood-ing, have seen increased use of coastal defencemeasures to protect human interests. Ironic-ally, this policy now threatens to increase thevulnerability of habitats to loss through coastalsqueeze. In the UK there is not one estuary thatwill not be vulnerable to coastal squeeze undersea-level rise. This is because they either haveextensive defences or they are rock-bound (ria)type systems. This has major implications forintertidal mudflats and salt marshes. Similarly,the coastlines of many deltas, and the margins of their major distributary channels, are oftenheavily defended and are thus also susceptible to squeeze.

Impacts of sea-level rise can be seen over arange of scales. On a large scale, the increasedloss of fringing salt marshes and greater pene-tration of the salt wedge further upstream are all ways of affecting the ecological and mor-phological structure of the system as estuaries widen and deepen. It is also feasible that theincreasing volume of sea water entering on theflood tide may alter flood/ebb dynamics andaffect sediment depositional patterns. Further-more, increased volumes of sea water withinestuaries may impede freshwater flow for longerperiods, leading to the possibility that in somelarge systems, where the volume of sea water is great, there could be an increased risk of riverine flooding further into the catchment(Bird 1993). Other, more subtle changes can beequally important. Sea-level rise has causedsalinity increases in some estuaries, which haveled to faunal and floral changes. In Louisiana,for example, salt marshes have replaced marginalreed swamps as water salinity and frequency of inundation have increased (de Sylva 1986),and in the coastal lagoons of the Nile delta, suchas Lake Burullus (see Case Study 7.2), increas-ing salinity of surface and groundwater hasaffected fish populations (Bird 1993). Similarsalinity increases also occur in coastal ground-water aquifers, thus posing an increased threatto drinking water supplies.

ESC07 9/7/06 3:19 PM Page 260


With respect to deltas, continued sea-level risemay have an impact on the seaward edge, in thatthe sea will start to dominate over the processesof river discharge and fluvial sediment input. Itis even argued that in many situations, the pro-gradation of deltas will cease (Bird 1993) andthat the low-lying lands of the delta will be moreprone to flooding as a result of storm-wave over-topping and increased brackish water penetra-tion up channels. Bird (1993) argues that a risein sea-level of 1 m would cause the submergenceof most of the seaward parts of the Nile delta,with the coastline moving several kilometresinland. Similarly, Milliman et al. (1989) suggestthat much of the delta region of Bangladeshwould be lost. Clearly, such change is dramaticand difficult to prevent. Defence provision is oneresponse, but to cope with a sea-level rise of 1 m,and given that storm-wave height would alsoincrease correspondingly with respect to currentlevels, the potential cost of the hard engineeringnecessary would be extreme.

On a smaller, but no less significant level, the survival of vegetation communities facedwith sea-level rise is linked to two factors. First, the issue of coastal squeeze can cause marshes to experience lateral retreat and loss of highermarsh communities (see Fig. 7.13). Second is the threat of drowning. For a marsh surface tosurvive in the face of sea-level rise, it needs togrow vertically and keep pace with the deepen-ing sea-level. Allen (1997, 2000) developed asimple model for quantifying this process. Giventhat the vegetation surface receives sediment intwo forms, minerogenic and organic (see Allen(2000) for detailed review), then the total amountof material received has to exceed the amount ofsea-level rise. Hence

ΔE = ΔSmin + ΔSorg − ΔM − ΔP

where ΔE is the net change in marsh surface elevation, ΔSmin is the thickness of minerogenicsediment added to the marsh surface, ΔSorg is the thickness of organic sediment added to themarsh surface, ΔM is the change in relative sea-level and ΔP is the height change resultingfrom compaction. If ΔE is negative, then the rate

of sea-level rise is greater than vertical marshgrowth, and the marsh is in danger of drowning.Conversely, if ΔE is positive, then the verticalgrowth is out-pacing sea-level rise. Althoughsuch rates can be measured for a contemporarysystem, prediction of the likelihood of futuremarsh survival is difficult. This is because thereare many unknowns with respect to estuarineresponse to progressive sea-level rise, particu-larly in respect of the sediment budget andhydrodynamics of the system.

Another significant aspect of sea-level rise isthat the associated increase in marsh erosionand increased height and salinity of the watertable will accelerate the reworking of a range ofenvironmental pollutants. Case Study 7.3 detailedthis idea from an erosion point of view, butchemical stability also becomes a factor whenground saturation and/or salinity increase. Thekey question to ask is where will this materialgo? Increased sediment loads due to marsh andmudflat erosion may be washed out to sea, ormay remain in the estuary, depending on tidalsymmetry. Either way, this may cause futuremanagement problems.

7.6.2 Impacts of increased anthropogenicdisturbance

Coupled with sea-level rise, humans still seeestuaries as areas of cheap land for development.Although with the advent of estuary manage-ment plans, development has become much morerestricted, there are still future issues that per-tain to this area of estuarine and delta impact.The growing leisure industry is imposing greaterdemands for marinas and sailing centres. As a1993 English Nature report stated, there was notone major estuarine system in England that didnot either have, or have plans for, a major marinadevelopment (Pye & French 1993). Similarly,the growth in ship size means that capital dredg-ing needs to be carried out in many ports andharbours in order to develop berthing capacity.Again, this represents further capacity to changetidal and current activity. Furthermore, as popu-lation grows, so the demand for fresh water willsee more dams being constructed in catchments.

ESC07 9/7/06 3:19 PM Page 261

262 PETER FRENCH

In Cyprus, much of the coastline relies on fluvialsediment supply for its main source of sediment.Owing to increased population and tourist num-bers, freshwater storage capacity increased from6 million cubic metres to 297 million cubic metresbetween 1960 and 1988 (Smith & Marchand2001). This has been achieved through increasedreservoir construction and the building of dams.The result is that little sediment now reaches the island’s estuaries, resulting in beach erosionalong some coasts of up to 0.5 m yr−1, ironicallydestroying the resource necessary for sustain-ing the tourist industry. The significance of thisincreased demand for fresh water cannot beunderestimated. Aquifers and rivers are the primesources of such water and increased extractionin both, but particularly rivers, will have keyimpacts in estuaries and deltas. The Nile delta(Case Study 7.2) is a good example, whereimportant lessons should be learnt.

One further future issue to threaten estuariesin particular lies in the method being increas-ingly used to mediate against intertidal loss as a result of sea-level rise. Managed realignment is increasingly being seen as a way to increasethe intertidal volume of an estuary, yet there are many unknowns about this technique thatneed to be investigated. As a comparatively newapproach to coastal management, its increasedusage has occurred against a background of littlefirm knowledge of the longer-term impacts anda range of issues have arisen from early uses ofthe method. Most notable appears to be the roleof pre-realignment vegetation on a site. Becausevegetation is effective in enhancing the trappingof sediment on developing marsh surfaces, itspresence can be seen as beneficial (Stoddart et al.1989; Fig. 7.6). In some examples of realign-ment, however, vegetation has not served thisfunction but has decayed following rapid burial(Macleod et al. 1999). This decay has led to the generation of anoxic conditions, which haveactually delayed marsh initiation. Other factorsinclude: larger scale changes in soil properties,often governed by what has happened in termsof improvement during its agricultural history;the change in groundwater dynamics, leading

to the release of stored contaminants; and gen-eral changes in ground water levels and salinity(Boorman & Hazeldon 1995; Crooks et al. 2002;Blackwell et al. 2004). Overall, the message hereis that although realignment presents a logicaland sound approach to increasing estuarinemarsh loss, there are important lessons still to belearnt. Some of these can be informed fromexisting schemes, and some from looking at historic storm-breached sites. Overall, however,there are still major questions to be answered,with much of the required knowledge having to come from existing realignment sites.

Much of what has been written has been pre-dicated by a presumption that scientists under-stand deltas and estuaries. In some areas ofknowledge, this is the case, but in others, theability to predict what may happen in the futureis based on a poor understanding of the principlesand functioning of these highly dynamic systems.The bottom line is that although we can measureand attempt to understand how estuaries worknow, our ability to predict change is still restricted.Perhaps the key issue here is that estuaries anddeltas are complex systems that contain a rangeof linking processes combining aspects of geo-morphology, sedimentology and ecology, as wellas human influences. One essential issue for thefuture, therefore, is to manage the whole systemas an entity with a range of component parts,and not to manage the component parts individ-ually. The ecosystem approach to managementis a philosophy that is gaining popularity acrossa range of systems. By combining populationdynamics with nutrient and sediment cycling,ecological production, sediment and water move-ments, and anthropogenic usage, managementcan be made much more effective. Mann (2000),however, reported that at the time of his writingefforts to produce predictive models of coastalecosystems, such as estuarine and delta systems,have had limited success. One of the key issueshere is the lack of adequate data sets to allowmodels of the complexity of estuarine interac-tions to be developed, and a second is to developsuch a model that can cope with the spatial andtemporal dynamics necessary.

ESC07 9/7/06 3:19 PM Page 262

8.1 INTRODUCTION

This chapter addresses the sedimentology of the clastic depositional environments of openocean temperate coasts. The environments thatwill be considered are associated with beaches, barriers and barrier islands composed largely ofreworked terrigenous material. These environ-ments are commonly sandy, but reference willalso be made to coarser grained gravel-, cobble-and boulder-dominated systems. These types of environment are typical of the temperate latitudes. In the tropics they are often replacedby systems in which organic materials dominate(Chapter 9). Estuarine and deltaic environmentsare considered in Chapter 7.

The common characteristic of the environ-ments considered is the non-cohesive nature ofthe constituent clasts. The consequent ability forclasts to readily be reorganized into different geo-morphological forms in response to changes inenvironmental conditions is a key characteristic.It is commonly asserted that beaches, as looseaccumulations of small grains of sand or peb-bles, are among the most unlikely landforms to exist in the often-harsh conditions of opencoastal areas (Pethick 1984). The very ability ofthese accumulations to change shape and absorbexcess energy in response to storms is central totheir ability to survive when more solid struc-tures, such as seawalls, can be destroyed underthe same conditions.

Thus open coast sedimentary environments,composed of a range of non-cohesive clast sizes,under a variety of energy conditions, take manygeomorphological forms. The controls on their

8Temperate coastal environments

Andrew Cooper

spatial distribution, the forms that they take andthe controls on their development are outlinedin this chapter. Morphological change over timeis discussed in order to provide an appreciationof the control of changing environmental condi-tions on coastal morphology. The interaction ofhumans with these dynamic coastal systems addsan additional layer of complexity to the dynam-ics and future evolution of temperate coasts,and thus past human interventions and futuremanagement options are assessed.

8.2 NATURE AND SIGNIFICANCE OF TEMPERATE COASTS

Temperate coasts (Fig. 8.1) are widely distributedin the Northern Hemisphere on Atlantic andPacific coasts of North America, Europe andNorth Africa, and the Far East. In the SouthernHemisphere they are limited to the Argentineanand Chilean coasts of South America, SouthernAfrica, and the southern coasts of Australia andNew Zealand. They are bounded toward thepoles by coasts with sea ice, and toward theequator by tropical coasts. Temperate coasts canbe crudely categorized into open ocean coasts(swell- or storm-wave dominated – see section8.2.3), sheltered seas and areas subject to tropi-cal storms (Fig. 8.1).

Sedimentary environments considered hereare beaches, barriers, barrier islands and coastalsand dunes (Fig. 8.2). A beach is an accumula-tion of wave-deposited, non-cohesive sedimentthat typically spans the subtidal–supratidal inter-face. Sustained beach sedimentation can giverise to a barrier (Woodroffe 2002). Barriers are

ESC08 9/7/06 3:20 PM Page 263

264 ANDREW COOPER

large, shore-parallel accumulations of sediment(of which a beach forms the most seaward part),deposited under wave action and which separ-ate a low-lying area (marsh, lagoon or estuary)from the ocean. Barrier islands are barriers thatare surrounded on all sides by water. Coastaldunes develop in the supratidal zone by aeoliantransport and deposition of sand from adjacentbeaches.

The critical conditions for development ofthese environments are as follows:

1 adequate sediment supply;2 suitable accommodation space (shaped byantecedent coastal morphology);3 sufficient wave energy levels to move avail-able sediment.Sediment is derived from various sources andhas varying texture. There is a broad latitudinalcontrol in that gravel beaches are preferentiallydeveloped in glaciated or formerly glaciated(paraglacial) regions (Ballantyne 2002). Sedi-ment texture is an important constraint on

T R O P I C S

22.5°S

22.5°N

Sea ice

Sheltered sea

Storm-wave dominated

Swell-wave dominated

Tropical cyclone tracks

Fig. 8.1 Distribution of temperate coasts including wave heights and types. (After Davies 1980; Short 1999.)

Washover DunesBeach

Ebb tidaldeltaFlood

tidaldelta

LagoonMainland

Barrier

Shoreface

Marsh

Fig. 8.2 Diagram of main sedimentaryenvironments considered in this chapter.

ESC08 9/7/06 3:20 PM Page 264

TEMPERATE COASTAL ENVIRONMENTS 265

beach morphology and behaviour. Differing clastsizes influence packing, resistance to motion andporosity, all of which affect beach morphology.The accommodation space in which beachesform is controlled by the antecedent coastal geomorphology. Commonly beaches form inembayments, but the geomorphology of theshoreline is an important factor in their distribu-tion. If too steeply dipping, sediment accumula-tion may not break the water surface and nobeach will form. Sufficient wave energy to trans-port available sediment sources is an obviousrequirement – insufficient energy will result in avariety of unmodified terrestrial deposits ratherthan beaches. The most important attribute ofthe environments described here is their abilityto change morphology as energy levels changeand to seek equilibrium with environmentalconditions. This ability to accommodate changerenders them important buffers against high-energy storm waves.

At a global scale, there is much variability inwave energy levels (Fig. 8.1). Swell waves onoceanic coasts are often fully refracted and arriveparallel to the shoreline, whereas locally gener-ated sea waves in sheltered environments mayarrive at an angle to the coast. Paradoxically,these lower energy waves may be more efficientin effecting net sediment transport throughlongshore drift. Tidal range exerts a secondaryinfluence on beach morphology as it mediatesthe vertical and horizontal area in which waveenergy is expended. Higher tidal range tends toreduce the relative influence of waves on beachgeomorphology.

The defensive significance of such malleablenatural environments to human developmentsat the coast has not been uniformly appreciated.There has been much historical degradation ofbeaches and dunes and in some localities this isongoing. The recreational and landscape attributesof temperate coasts combine to draw increasingnumbers of humans to the coast. The consequentinfrastructural development is often poorly loc-ated and impedes the ability of natural beachsystems to respond to changing conditions.

The environmental significance of temper-ate coastal sedimentary environments can be

expressed or considered from several, often com-peting, human perspectives (physical, ecologicaland economic). In a physical sense, the environ-ments provide a dynamic natural system capableof responding to and absorbing high levels ofmarine energy. These environments are oftenregarded as hazardous to navigation. From an ecological perspective, beaches have beenrecognized, over the past 20 years in particular,as important ecosystems (Brown & McLachlan2002). The interstitial fauna and flora play a rolein the ecology and sedimentology of adjacentinshore and dune ecosystems as do their distinc-tive vegetation and related fauna.

The economic significance of beaches is evid-ent in the extent of seafront development andthe vast numbers of visitors to resort beaches.As a coastal defence against wave action, beachesare also economic assets. Beaches too provideaccess to the coast and they have also been seen as ready sources of aggregate (Carter et al.1991). The human values attached to beaches areclearly not always compatible with each otheror with their ecological and physical value. Thenature of human pressure on beach and duneenvironments changes with time (Nordstrom2000). Much of this is technology-dependent.For example, widespread kelp collection in thenorth-west British Isles during the Napoleonicwars for iodine extraction undoubtedly affectedthe development of drift lines, so necessary fordune development. Similarly, changes in farm-ing practice have controlled the patterns ofsmall-scale sand removal from beaches in Ireland(Carter et al. 1991).

Human pressure on the coast takes many formsand the continuing migration to the coast is setto increase those pressures. Cohen et al. (1997)estimated that over 2 billion people (37% of the global population) live within 100 km of acoastline. Dramatic increases have been noted inthe temperate regions of the Mediterranean andthe USA ocean coasts. In the Mediterranean, thecoastal population was estimated at 146 millionin 1990 and the urban coastal population aloneis projected to rise to 176 million by 2025, withan additional 350 million tourists (Hinrichsen1998). In the USA, 55–60% of the population

ESC08 9/7/06 3:20 PM Page 265

266 ANDREW COOPER

live in the coastal counties of the Atlantic andPacific coastlines and coastal population densityrose from 275 to 400 people per square kilometrebetween 1960 and 1990 (Hinrichsen 1998).

8.2.1 Environmental sedimentology of temperate coasts

The world’s temperate coasts are among themost densely populated and developed in theworld. As a consequence, large stretches of temperate coastlines have been subject to inten-sive anthropogenic modification with attendant sedimentological impacts. In many cases theseinvolve direct impacts on the sediment volume(e.g. dune removal, sand mining) or sedimentaryprocesses (e.g. jetty construction interferingwith longshore drift). In other instances, in-direct influences result from, for example, thearmouring of coastal cliffs, which formerly supplied sediment to adjacent beaches. In allcases, human development of the coastline hasthe potential to foreclose future managementoptions. Intensive infrastructural developmentat the coast almost inevitably leads to the desireto armour the shoreline in the face of coastalerosion. Armouring fixes the shoreline position,thus setting a human limit to the acceptableextent of shoreline processes. This constrainsthe ability of the beach system to operate natu-rally and passes the fate of the beach into the human decision-making sphere. This hasgiven rise to the study of ‘developed coasts’ as a distinctive field of investigation (Nordstrom 2000). The distinction is somewhat blurred as even remote, rural beaches often exhibitsignificant human alteration and impact (Poweret al. 2000).

8.2.2 Overview of issues, processes and problems

There are many contemporary managementissues in the temperate coastal zone, most ofwhich involve conflicts between human utiliza-tion and natural processes of change. Even in an environment with predictable dynamics anda fully understood coastal system, the issueswould still be difficult to resolve because of the

human value system, politics and economics thatinfluence human responses to pressure (Pilkey& Dixon 1996). In reality, the dynamics of the coastline are unpredictable (storms occurchaotically) and the coastal system is not wellunderstood (the complexities and feedbacks in acoastal system are largely unresolved). Further-more, sea-level, the plane at which coastaldynamics operate, is changing on a global scale,and global climate change may be altering thefrequency and intensity of storms. This is mani-fest in many parts of the world as a propensitytoward coastal recession (Bird 1985). Added tothis is ongoing deliberate and accidental modifica-tion of sedimentary processes through humanintervention.

8.3 SEDIMENT SOURCES AND SEDIMENT

ACCUMULATION PROCESSES

8.3.1 Sources and characteristics of sediments

Temperate coastal environments receive sedi-ment from various sources and inevitably beach morphology is strongly influenced by the nature and volume of available sediment(Fig. 8.3). The main sources are rivers, the con-tinental shelf, coastal erosion and redistribution(see Chapter 1) and there are important regionaldifferences in the type of sediment supply. In thewestern USA fluvial sediment supply is domin-ant because of the steep hinterland gradient,whereas on the eastern coast, continental shelfsources predominate.

Fluvial sources are a key component of tem-perate sediment supply, especially adjacent tolarge rivers or rivers that drain steep hinter-lands. Fluvial sediments may be reworked withina delta complex, often to form barrier islands, for example the Nile delta (Pilkey 2003). Inother instances, sediment is transported fromthe river mouth to adjacent coastal depocentres(Cooper et al. 1999). Fluvial sediment supply is often episodic, particularly where a seasonaldischarge regime exists (see Chapter 3). Thuscoastal sediment supply may fluctuate betweenperiods of high fluvial supply followed by intervals

ESC08 9/7/06 3:20 PM Page 266


during which sediment is reworked by coastalprocesses (Cooper 1993).

The continental shelf is a vital sediment sourcefor many temperate coastlines. Here relict ter-restrial, marine or coastal sediments from lowersea-levels or past glaciations may be reworkedlandward by contemporary wave action. Off-shore sediment abundance was accompanied bymillennial-scale coastal progradation in south-east Australia (Thom 1984) and decadal-scalebeach ridge accumulation in south-east Spain(Goy et al. 2003). In conditions of sedimentscarcity, wave action may, however, erode thesea-floor to transport older deposits to the shore-line. In North Carolina, grains eroded fromunderlying Tertiary lithologies are a componentof contemporary beach sediment (Pilkey et al.1998). Sediment transport on the shelf variesaccording to wave conditions and during storm-wave action extends to deeper water. Thus theshelf sediment supply may also be temporallyvariable (see Chapter 10).

The coastline itself is a source of sediment to adjacent beaches. Often, longshore drift ofsand from adjacent beaches and cliffs is a majorelement in the sediment supply, although this isdiminished in strongly embayed coasts. Erosionof relict glacial deposits yields abundant graveland boulder beaches in the higher latitudes(Davies 1980). Elsewhere, the lithology and texture of coastal outcrops strongly influencethe nature of eroded clasts that supply beaches.In southern England, for example, the prefer-ential preservation of chert (flint) clasts erodedfrom less resistant limestone (chalk) is marked(Carr 1969). The supply of sediment is not con-stant and is controlled to a large extent by patternsof slope failure on coastal cliffs, and the capacityof waves to transport sediment. Slope failuredepends on a range of factors including theporosity and permeability of the rock, stratifica-tion, rainfall intensity and the extent of waveundercutting (Emery & Kuhn 1982). Often sedi-ment input is in the form of large-scale landslides

Cliff erosion

River input

Delta

Dam

Drift

Drift

Lagoon

Nourishment

Beachridge plain

Urban

Transgressivedune

Longshore

EstuaryRiverCliff

DuneInlet

Nourishment

Longshore

Offshoreloss/gain

Extraction

In situcoastal

production

Onshore/

Offshore

Inlet exchange

Groynes

Shor

efac

e

Fig. 8.3 Diagrammatic representation of coastal sedimentbudget showing main sources and exchanges both natural andanthropogenic.

ESC08 9/7/06 3:20 PM Page 267

268 ANDREW COOPER

that occur episodically. Brunsden & Moore(1999) documented a landslide on the DorsetCoast (England) that blocked alongshore sedi-ment supply to adjacent beaches.

Estuaries, depending on their hydrodynamics,may be either sediment sources or sinks in thecoastal zone (Cooper 2001a) (see Chapter 7).Flood currents may produce landward transferof bedload sediments (sand and gravel), whereasebb-currents, often augmented by river discharge,induce their seaward transport (see Chapter 1).Reworking of sediments among different ele-ments of the coastal system produces local sourcesand sinks of sediment in response to changingdynamics. Cycling of sediment between tidaldeltas, inlets and beach-dune systems exem-plifies the process (Fig. 8.4). Temporal lags insediment transfer rates between the variousparts of the inlet system may lead to periodicerosion and accretion on beaches, and episodicevents such as storms strongly influence rates ofsediment transfer (Morton et al. 1995).

The contribution of organic materials in tem-perate latitude beach systems is typically muchless important than in the tropics. At sites of low

terrigenous sediment supply, however, carbonatematerials may dominate the coastal sediment.Barnhardt & Kelley (1995) documented a beachin Maine where sediment was derived from off-shore mollusc fragments. In western Ireland,carbonate-rich beaches composed of foraminiferaand coralline algae are locally abundant (Bosence1980). Less persistent, but volumetrically sub-stantial organic components of temperate beachesinclude woody debris and marine algae. Althoughthese components eventually decay they provideimportant nutrient sources, and foci for sub-sequent aeolian accumulation. In areas wherebeaches are underlain by back-barrier sediments,eroded peat or mud may form clasts on contem-porary beaches.

8.3.2 Temperate coastal depositional landforms

A suite of coastal landforms develops in tem-perate latitudes in response to sediment supplyand a suitable sediment accumulation zone ortrap. These landforms (Fig. 8.5) at the macro-scale include barriers (Fig. 8.6a), barrier islands(Fig. 8.6c), and mainland-attached beaches in

Ebb spit

Sediment transport paths

Lagoon Open ocean

Flood-tidal delta

Ebb-tidal deltaEbbshield

Flood channel

Dominant waves

Dominant waves

Flood rampMain ebbchannel Terminal lobe

Inletsediment bypassing

Overwash

Inlet

throat

Rec

urve

d sp

it

Barrier island

Marginal flood channel

SwashbarM

argi

nal f

lood

cha

nnel

Fig. 8.4 Main geomorphological features and sediment transfer pathways at a tidal inlet. (After Hayes 1991.)

ESC08 9/7/06 3:20 PM Page 268


the form of cuspate forelands (Fig. 8.6b) andspits (Fig. 8.6d). Barrier islands are surroundedon all sides by environments that are regularlysubmerged. Thus a barrier island is fronted bythe open ocean and backed by a sheltered lagoon(which may be marsh-filled). They are boundedat either end by inlets that permit exchange ofwater and sediment between the back-barrierenvironment and the open sea.

Barriers enclose a marsh, lagoon or estuary and,unlike barrier islands, are mainland-attached.An inlet may or may not be present, throughwhich water flows between the back-barrier system and the ocean. In some instances theseinlets are ephemeral or seasonal features depend-ent on freshwater flow (Cooper 2001b). Inletson both barriers and barrier islands may or may not contain flood- and ebb-tidal deltas. Thepresence and morphology of tidal deltas arecontrolled by the relative strengths of waves and

Fig. 8.5 Temperate depositional landforms (headland–embayment beach, spit, tombolo, salient, cuspate foreland,barrier, barrier island).

tidal currents, coupled with the requirement fora suitable accommodation space in which to form(Oertel 1985).

Mainland-attached beaches are backed bydry land rather than water. These beaches (andsome barriers) may be strongly influenced by the onshore and nearshore topography. A number of distinctive beach planform shapesdevelop in response. These include headland–embayment systems and cuspate forms (salients,forelands and tombolos) (Sanderson & Elliot1996).

Coastal dunes constitute the aeolian com-ponent of many temperate beach systems andexhibit a wide range of morphologies (Fig. 8.7).In temperate regions many coastal dunes arevegetated. The dunes begin as accumulationsaround debris on the beach, typically drift linesof seaweed and other material. Initial coloniza-tion by strand-line vegetation partly stabilizesthese dunes and may permit continued accretionas wind-blown sand is trapped by the vegeta-tion. Dunes comprise varying areas of vegetatedand unvegetated sand and are intimately linkedto their source beach areas (Sherman & Bauer1993).

On prograding systems with abundant sedi-ment supply and strong vegetation growth,shore-parallel lines of foredune ridges mayform. Less vigorous vegetation growth leads to a less organized, hummocky appearance.Very dense and vigorous vegetation can lead to effective sediment trapping and vertical dune growth. A steep nearshore slope probablyaccentuates this form of dune growth by pre-cluding seaward progradation. In the absence of vegetation a transgressive sand sheet mayform, which migrates landward across pre-existing features. Instability and vegetationdamage may lead to blowout formation on vegetated dunes and, if persistent, will cause thedevelopment of parabolic dunes that migratelandward through the dune system. As a result,dunes must be regarded not only in terms ofinteractions between the beach and dune but as dynamic systems in their own right in whichmorphological change and sediment transfermay proceed intermittently.

LagoonBarrier island

Headland

Barrier

Tidal delta

Headland

tomboloBedrock

Bay beach f

Spit

Lagoon

Dune

Erodingcliff

SpitDeepembayment

Estuarydelta

Salient

Bedrock island

Tidal delta

Sandy delta

Beachridge plain

Coastalplain

ESC08 9/7/06 3:20 PM Page 269

270 ANDREW COOPER

Fig. 8.6 Depositional landforms: (a) Chesil Beach, a barrier (arrowed); (b) Dungeness, a cuspate foreland; (c) The Frisian Islands, abarrier island chain; (d) Spurn Head Spit at the Humber Estuary mouth. (Images from NASA Earth Science Applications Directorate(https://zulu.ssc.nasa.gov/mrsid/).)

Sand transport

(a) Foredune ridges (b) Hummocky or transverse dunes

(c) Transgressive sand sheet (d) Single accreted foredune

(e) Parabolic dunes and blowouts

Sand supplied to beachand dunes, incompletevegetation cover

Beach retreat.Sand drift in transgressivesand sheet

Beach retreat.Blowouts result inparabolic dune migration

Rapid sandsupply trapped on vegetated fore dune

t2t1

t2 t1

t4t3

t2

t1

t4t3

t1t2

Fig. 8.7 Typical dunemorphologies. On gentle gradientbasement slopes (a), shore-parallelbeach ridges form. Foredune ridgesmay develop on each beachridge toform a sequence of dune-toppedridges. If vegetation is patchy, thedune topography is more irregular(b) and if sparse or absent, atransgressive sand sheet (c) mayform. On steep basement slopeswith good vegetation growth,vertical accretion may give rise to ahigh, single dune ridge (d).Blowouts are associated with breaksin vegetation cover and whichinvolve cannibalization of the dunesand by ongoing wind action (e). In time parabolic dunes may evolvethat transport sediment landwardthrough the dune system Temporalevolution is denoted by t1–t4. (After Woodroffe 2002.)

ESC08 9/7/06 3:20 PM Page 270


8.3.3 Controls on sediment accumulation andtransport

Sediment transport in the nearshore zone is ahighly complex phenomenon that occurs througha variety of interlinked processes operating atspatially and temporally variable intensity. Asynopsis is presented here; for a more thoroughreview of sediment processes in the nearshorethe reader is referred to Komar (1998) orWoodroffe (2002). The dominant force in sedi-ment transport and geomorphological change in the coastal zone is wave action. Waves areformed by winds blowing over a water surface.They are typically described in terms of wavelength, height and period. Once formed, wavesradiate out from the generation area and thewave form propagates across the water surface.Typically, a wide range of wave sizes (defined bywavelength and period) are produced by windsat the source area. The waves formed directly inthe area of generation tend to be of various sizesand the water is very ‘choppy’. Larger waves are produced by winds of greater speed andduration, and greater fetch distances. As wavestravel further from the generation area theybecome sorted and amalgamated. Longer wavestravel faster and arrive at distant coasts first as a regularly spaced set of large waves (termedswell). Oceans have large fetches and hence largewaves are produced. Coasts facing open oceansare thus typically dominated by such conditions.In semi-enclosed seas, most waves are generatedby more proximal winds blowing across restrictedfetches and hence the waves are smaller andmore variable in their dimensions.

As waves move from deep to shallow water,friction with the sea-floor retards the wave,which maintains its energy by steepening. Waveswith longer wavelengths ‘feel’ the sea-floor first and are slowed comparatively far from the shore. The point at which significant wave–sediment interaction begins is known as wavebase (see Chapter 1). It occurs at variable depthdepending on the length of the waves. Wavesapproaching a shoreline obliquely thus changethe orientation of their crests in response to bottom bathymetry. This process by which the

wave crests are reorientated is known as re-fraction and it is important in distributing waveenergy along a shoreline. A related process ofdiffraction, by which wave energy is transferredlaterally along the crest, also occurs as wave crestsbend. Because swell waves have longer wave-lengths they interact with the sea-bed further off-shore and hence arrive more fully refracted thanshort wavelength waves. Full refraction tends toarrange wave crests parallel to the shoreline andto equally distribute energy alongshore. Shortperiod, locally generated waves exhibit greaterenergy differences alongshore due to incompleterefraction.

Waves undergo further modifications as theyapproach the shore and a series of zones aredefined according to wave behaviour near thecoast (Fig. 8.8). As waves interact with the sea-bedand slow, they become higher. The zone in whichthis takes place is known as the shoaling zone.Wave steepening may lead to instability as thewave becomes too high relative to its length. Itthen ‘breaks’ as gravity collapses the waveform.A range of breaking criteria has been determinedempirically (Komar 1998), and several breakertypes are recognized in a continuum includingsurging, collapsing, plunging and spilling breakersrelated to the mode of energy release.

Landward of the breaker zone, wave energycontinues to propagate shoreward in what isknown as the surf zone (Fig. 8.8). This is a zoneof often intense turbulence as waves reform,break again and generate secondary wave andcurrent motions. Much wave energy is dissipatedin the surf zone through breaking, turbulenceand sediment transport. The surf zone is vari-able in width. Wide surf zones are associatedwith large waves, which dissipate their energythrough a series of spilling breakers. Lowerenergy beaches have narrow surf zones in whichwave energy is dissipated through breaking closeto the shore. Reorganization of wave energy can lead to generation of secondary wave forms(edge waves) that have longer wave periods than incident waves. These have crests that areoblique or perpendicular to the shoreline. Theycan be stationary or may propagate along theshore and give rise to widely spaced inequalities

ESC08 9/7/06 3:20 PM Page 271

272 ANDREW COOPER

in water level, often observed as regularly spaced,alongshore peaks in swash. In addition, second-ary currents may be generated due to inequalitiesin the water surface elevation and energy density.These currents may be arranged into circulatorycells that redistribute water through shore-normal (rip) and/or shore-parallel (longshore)currents (Fig. 8.9).

Energy that propagates to the shore throughthe surf zone to the beach may run up the beachfor some distance as swash. The elevation towhich swash rises is dependent on wave height,the porosity of the beach material and the extentof saturation at any given time. Excess waterthat does not infiltrate the beach runs backdown the beach as backwash. The high porosityof gravel beaches reduces or eliminates back-wash. Swash is normally the zone in which final dissipation of wave energy takes place. Onsteep beaches, however, excess energy may bereflected seaward, or, if the swash reaches thebeach berm, it may flow in a landward directionas overwash.

Waves are thus responsible for a variety of fluidmotions and potential sediment-transportingmechanisms on temperate coasts. Commonly,for ease of understanding these are divided intocross-shore and longshore transport mechan-isms, although in reality the system is fully

Breaker zoneSurf zone

Intertidal

zone

LWM

HWM

Berm

Foredune

Bar

Beachface

MSL

Backshorezone

Nearshore zoneor shoreface

Offshorezone

SwashzoneSand

dune

Tidalrange

Fig. 8.8 Nearshore wave zones and profile morphology: HWM, high water mark; LWM, low water mark; MSL, mean sea level. In theoffshore zone waves begin to interact with the sea-bed, and increase in height as the water depth diminishes. Waves break, generatingbars as they do so, in the breaker zone and propagate across the surf zone toward the shoreline. The final wave energy dissipation occursin the swash zone. Overwashing occurs when waves overtop the berm and flow landward across the backshore. Aeolian transport ofsand from the beach may produce dunes at the rear of the beach.

(a)

(b)

(c)

Beach

Beach

Breaker zone

Ripcurrent

Ripcurrent

ReturnWave crest

Longshore currentWave crest

Beach

Reflection Reflection

EdgewaveEdge wave refraction

Wave crest

Fig. 8.9 Nearshore circulations in the surf zone take the form ofcirculation systems (a) comprising rip currents and alongshorecurrents, (b) shore currents generated by oblique incident wavesand (c) edge waves produced by reflection of incident waveenergy from steep beaches and subsequent refraction backtowards the shoreline.

ESC08 9/7/06 3:20 PM Page 272


three-dimensional. Longshore sediment trans-port takes place largely via waves that approachthe shoreline obliquely. These move sediment in the direction of the dominant wave approach.The process is, however, highly variable. Largerwaves interact with the sea-bed further offshoreand extend the active transport zone. Cross-shore sediment transport involves the onshoreor offshore movement of sediment in responseto changing wave character and bed slope.

The main role of tidal water level variation on sedimentation on beaches is to mediate theplane at which wave processes operate andhence coasts are commonly categorized on thebasis of tidal range (see Chapter 7). In areas of higher tidal range, wave energy is spread over a greater vertical range (and hence lateralextent) and is less effective than in areas of lowtidal range. Low tidal range has often been asso-ciated with wave dominance on coasts (Davis & Hayes 1984), however, Anthony & Orford(2002) have recently drawn attention to the features of mixed-energy coasts with high levelsof both wave and tidal control. Each of thewave-generated energy fluxes outlined abovehas the potential to transport and reorganize thenon-cohesive sediments of beaches. The nature

and relative strength of these water motions will influence the direction and magnitude ofsediment transport, as will the dimensions andtexture of the beach sediment. It is also import-ant to recognize the feedback relationships thatexist between morphology, fluid dynamics andsediment transport (Fig. 8.10).

Although actual sediment transport takes the form of individual grains moving in suspen-sion, saltation or bedload (see Chapter 1), theintegrated transport volumes are of interest inunderstanding beach behaviour. The link betweencoastal landform and the formative processes hasgiven rise to the field of study known as mor-phodynamics in which the relationship betweendynamic forcing and morphological response is considered (Carter & Woodroffe 1994). Thetypical approach to sediment transport at thecoast is to consider longshore and cross-shorecomponents separately, although both operatetogether. Movement of beach sediment in a shore-normal direction is largely driven by inequalitiesin the landward and seaward velocity and dura-tion of wave-induced currents. The dominantmechanism in longshore transport is movementof sediment by longshore-directed currents gener-ated by oblique waves (USACE 2003). As most

Process elements Response elements

Feedback

Energy factorsWaves: height, period,angle of approach.Tides: range, diurnalpattern, stage.Currents: velocity,direction.Wind on backshore, velocity, direction

Material factorsMean grain diameter,sorting, mineralcomposition, moisturecontent, stratification.

Shore geometryStraight, curved,bottom slope, gentle,steep.

Beach geometry

Foreshore slope, width, heightof berm, backshore width.

Beach materialsMean grain diameter, sorting,mineral composition, moisture content, stratification

Fig. 8.10 Conceptual model of the coastal systemillustrating process elements (energy, materials andgeometry) and response elements (geomorphologyand texture) and feedback between the two.Variability in all factors and the presence offeedback, render quantification of nearshoreprocesses difficult. (After Krumbein 1963.)

ESC08 9/7/06 3:20 PM Page 273

274 ANDREW COOPER

wave energy dissipation takes place in the surfzone, this is a zone of high sediment transportpotential.

The processes that affect coastal sand dunesare quite distinctive as they rely not upon waveaction, but wind. The presence of coastal dunesrequires three criteria to be met. First, theremust be a supply of sand. This is usually theadjacent beach and beach supply is enhanced if the sand is dry and transport is not impededby coarse fractions. Second, the wind velocitymust be sufficiently strong to entrain grains andmust have a significant onshore component. The third requirement is an interruption in the airflow such that velocities drop and sand is deposited. This is commonly provided by aphysical obstacle on the beach, but may also berelated to topographic features or vegetation,which cause airflow to diverge. Potential con-straints on aeolian transport on beaches includefetch distance, moisture, diagenesis (e.g. salcretes),surface roughness and armouring (Sherman &Bauer 1993).

8.3.4 Coastal morphology

Just as sediment transport is commonly con-sidered in longshore and cross-shore contexts,the same is true of beach morphology. Below,the morphology of beaches is considered first ina cross-shore (profile) context and second froman alongshore (planform) perspective. In realitythe two are closely related in the fully three-dimensional beach morphology.

8.3.4.1 Beach profile morphology

The cross-shore geometry of a beach is stronglyinfluenced by its constituent grains. Boulderbeaches (clast diameter > 0.25 m) have low gradients (typically 6–14°). This is lower thanthe natural angle of repose and results from the fact that they are flattened by extremestorms and there is no mechanism to rebuildthem. Gravel beaches (clast diameter > 2 mm) in contrast tend to be steep (> 15°) because thelarge pore space promotes infiltration of waterrather than surface backwash. In contrast sand

beaches are gentler in gradient. In general, sandbeaches assume lower gradients (usually < 10°)with higher wave energy; the reduced gradient is associated with dissipation of wave energy.Paradoxically, this may result in lower residualenergy at the coast than on steeper sand beachesassociated with lower wave energy. In the lattercase, residual energy is reflected from the beach.The terms dissipative and reflective have thusbeen used to describe sand beaches in a numberof morphodynamic classifications. Beaches withmixed grain sizes tend to show some degree ofspatial segregation of the grains. Typically thistakes the form of a high tide, steep beach facetcomposed of coarse material, fronted by a low-gradient sand ‘apron’. Such beaches compriseboth dissipative and reflective elements.

Beaches show a range of features (Fig. 8.8)that are developed to greater or lesser extentsdepending on several factors. The beachface isthe zone in which waves are altered as theyapproach the shoreline. The berm is the limit towhich swash action typically extends, and is anaggradational feature. The back-beach typicallyslopes gently landward and is activated by over-wash processes. It also acts as the source zonefor dune sands. The beach may merge into adune system in which ephemeral dunes give wayto a foredune ridge that in turn may be backedby more stable dunes.

8.3.4.2 Coastal planforms

The processes of sedimentation outlined abovehave high potential to sort and arrange indi-vidual clasts into distinctive landforms. In mostcircumstances, there is a limited sediment sup-ply, and in all cases, a topographic control onthe distribution of sediments along the coast.Thus coastal sedimentary systems are discon-tinuous and form discrete systems. Althoughsome sedimentary systems are physiographic-ally distinct from a planform perspective (e.g.headland–embayment cell), others on linearclastic coasts are more difficult to compart-mentalize. The concept of coastal cells enablesidentification of semi-closed zones within whichmaterial fluxes may be quantified (even if in

ESC08 9/7/06 3:20 PM Page 274


broad terms). Such cells are commonly definedin terms of the two-dimensional horizontal planearrangement of sediments, known as the coastalplanform (Fig. 8.11). The simplified approachof May & Tanner (1973) in which longshoregradients in wave energy at the breakpoint are quantified, provides one approach to thedelimitation of cells for any given wave condi-tion. A number of equilibrium cell forms havebeen identified (Fig. 8.11) (Carter 1988).

In swash-aligned equilibrium, wave energy is distributed evenly along a shoreline such thatthere is no longshore wave energy gradient toproduce sediment transport. In a graded equi-librium, the beach sediments become sorted inresponse to longshore gradients in wave energy.The phenomenon is best seen on gravel or cobblebeaches and is manifest by alongshore gradientsin clast size such that the clasts at any given pointare too large to be transported by wave energyat that point. On coasts with dominantly obliquewave approach, drift-aligned equilibrium may beattained at any given point if inputs of sedimentfrom up-drift sources are matched by losses todownstream sinks. Particular types of equilib-rium forms (zeta and log-spiral bays) compriseboth swash-aligned and drift-aligned sectionsand occur where a headland obstructs longshoredrift incompletely (Woodroffe 2002).

8.4 PROCESSES AND IMPACTS OF NATURAL DISTURBANCE

AND ENVIRONMENTAL CHANGE

8.4.1 Temporal change in temperate beach anddune morphology

Beaches and dunes are characterized by a widerange of types, and grain size provides a first-order discrimination among boulder, gravel andsand beaches. Although these types form a con-tinuum there are distinctive modes of behaviourand response to environmental forcing for each.Boulder beaches (Oak 1984) are inactive undermost coastal conditions because of the largeclast sizes involved. Extreme (high magnitude,low frequency) events (e.g. storms and tsunami)thus create the only circumstances under whichthe beach morphology changes. The boulderbeach gradient at any time reflects a combina-tion of the intensity of the last dynamicallyeffective storm, as well as the cumulative workof a series of such storms. An interesting feed-back relationship thus exists in that the mor-phodynamic effectiveness of successive stormsof the same magnitude is reduced by the beachhaving moved closer to equilibrium with thatmagnitude of storm.

Gravel beaches have been shown, largely byvirtue of their porosity, which inhibits backwash,

(a) (b) (c) (d)

Fine

Coarse

Wavecrests

Longshoredrift

Flatbeach

Steepbeach

Fig. 8.11 Cell equilibrium types: (a) swash-aligned equilibrium where wave crests arrive at the shoreline with equal distribution ofenergy alongshore; (b) drift-aligned equilibrium where sediment flux equals the potential transport ability of waves; (c) gradedequilibrium where grain size at any point is too great for wave transport; (d) log spiral or zeta bay.

ESC08 9/7/06 3:20 PM Page 275

276 ANDREW COOPER

but also their comparative resistance to movementunder low wave energy, to exhibit landwardmigration via barrier overwashing. High poros-ity precludes a seaward return mechanism forbeach clasts and thus gravel beaches tend to besteep because of the higher angle of repose ofcoarse clasts. Sand beaches are the most widelydistributed and morphologically variable of beachtypes. For this reason they have been studiedmore intensively than other types. Attempts tosynthesize studies of beach morphology havebeen made, perhaps the best known resulting inthe concept of beach state models (Wright &Short 1984; Masselink & Short 1993). Thesestudies recognize spectra of beach types that

are arrayed on a dissipative (low-angle beachgradient) to reflective (steep gradient) continuum(Fig. 8.12).

Dissipative beaches are associated with longperiod, large waves with high incident energy.Energy is dissipated by wave breaking across awide, low-angle shoreface such that waves at theshoreline have little remaining energy. Becausesuch beaches are associated with long-periodwaves, these are often fully refracted by the timethey reach the shoreline and hence temporalmorphological variability on such beaches iscomparatively low. Reflective beaches at theopposite end of the continuum are associatedwith coarser sediment and/or smaller waves of

Cusps

Step

HTLT

0 100

0

−2

CuspsReflective beachface

Low tide terracewith rips

HT

LT

0 100 200

0

−2

Transition to tide-dominated tidal flats

CuspsSteep beachface HT

LT

0 100 200

0

−2

Steep beachfaceHT

LT

0 100 200

0

−2

Steep beachface HTLT

0 100 200

0

−2Bar 1

Bar 2Bar n

0 2 5

HT

LT

0 100 200 300

0

−2

−4

−6

Reflective high tide beach

Dissipative low tide terrace

(Cusps)

Steep beachface HT

LT

0 100 200 300

0

−2(Swash bar)

HT

LT

0 100 200 300

0

−2

−4

HT

LT

0 100 200 300 400 500 600

0

−2

−4

−6

157

31

0Dimensionless fall velocity W = Hb/wsT

Rel

ativ

e ti

dal

ran

ge

RT

R =

MS

R/H

b

Reflective Intermediate Dissipative

Reflective Barred Barred dissipative

Low tide terrace + rip Low tide bar/rip Non-barred dissipative

Low tide terrace Ultra-dissipative

Refective beachfaceDeep trough Pronounced

bar

Deep trough

Subdued multiple bar-trough morphology

Low tidetransverse bar and rip

morphology

Flat and featureless

Flat and featureless

Fig. 8.12 Conceptual beach state models for micro- to macrotidal beaches: MSR, mean spring tidal range; Hb, breaking wave height;Ws, sediment fall velocity; T, wave period. With increasing breaking wave height and/or decreasing grain size, beaches become gentler in gradient and more ‘dissipative’ of incident wave energy. The relationship of tidal range and breaking wave height (relative tidal range)mediates the vertical range over which wave energy is dissipated. Thus beaches with larger tidal range become progressively moredissipative. (After Masselink & Short 1993.)

ESC08 9/7/06 3:20 PM Page 276


shorter period. These beaches tend to permitpropagation of incident wave energy closer tothe shore and thus waves break closer to theshoreline with more residual energy than on dis-sipative beaches. Excess energy is reflected fromthe beach face, which tends to be steep. Wavesare often not fully refracted at the shoreline andare thus more likely to generate secondary wavemotions and currents. The presence of beachcusps, and high longshore sediment transportrates are physical manifestations of such condi-tions. Intermediate beaches between these twoextremes display both dissipative and reflectiveelements. On such beaches, a low-angle dissip-ative facet is backed by a steep, reflective facet,and wave energy is accommodated by a com-bination of dissipation and reflection.

Empirical data have been used to relate beachstate to various morphodynamic indices includ-ing Dean’s parameter (combining grain size,wave height and period) and surf scaling para-meter (combining nearshore slope, wave heightand period) (Wright & Short 1984; Masselink& Short 1993). These beach state models providea set of ‘expectation criteria’ for beach morpho-logy in a given dynamic setting. A number ofauthors (e.g. Hegge et al. 1996) have questionedthe universal applicability of beach state modelsand it is likely that factors such as sediment supply, underlying topographic control and vari-ations in wave climate may well create a widervariety of beach states than is accommodatedwithin existing models.

Within the beach state approach, temporalchanges in factors such as wave height arematched by corresponding changes in beachmorphology that take place in a predictablesequence. These changes are largely describedby variations in profile shape, driven by cross-shore sediment transport. With decreasing waveenergy, changes take the form of bar migrationonshore and eventual welding onto the beach-face. Increasing wave energy leads to a loweringof beach gradient and bar formation on theshoreface. Beaches with strongly seasonal waveclimate variation may traverse the full range of beach states (Shaw 1985), whereas othersremain within the same state perennially. There

are many beaches that do not appear to fit thebeach state models and the role of sedimenttransport thresholds remains a constraint onsuch models, whereby beach morphology mightreflect only higher energy events.

8.4.2 Planform variability

The planform morphology of beaches and barriers in relation to sediment supply andaccommodation space has been considered inthe previous section. Some aspects of the grossmorphology of beaches and barriers relate tospatial and temporal variations in dynamics. Themost commonly recognized control is the relativeimportance of wave and tidal processes. This wasidentified in a regional study of the eastern USA(Hayes 1979; Davis & Hayes 1984). In terms of barriers and beaches these studies identifiedwave-dominated and mixed wave- and tide-dominated systems. Wave-dominated barrierislands tended to have relatively few tidal inlets(a result of small tidal prisms), small ebb deltas(a result of wave reworking) and well-developedflood deltas. Mixed-energy coasts in contrasttended to have more frequent inlets (larger tidalprisms requiring more inlets for tidal exchange),larger ebb deltas (stronger tidal currents) andmore extensive marshes (a consequence of flood-tidal current deposition of fine sediments). Sucha division has not been recognized on gravelbarriers, perhaps because of their latitudinallyrestricted distribution.

Beaches and barriers are subject to manyadditional changes in their overall planform.Typically these changes are more gradual thanthe types of change envisaged in beach statemodels. Changes related to sea-level rise are discussed below. Here we are concerned withchanges that may be effected independently of sea-level change. Beach planform adjusts, as do profiles, to minimize variations in waveenergy. The planform adjustment is markedlyconstrained by the topographic setting of thebeach, and the sediment supply. Abundant sedi-ment supply leads to progradation (Fig. 8.13).Such progradation involves successive weldingof bars, which provide sediment supply for

ESC08 9/7/06 3:20 PM Page 277

278 ANDREW COOPER

embryo dune and then foredune growth. Withcontinuing sedimentation the foredune is increas-ingly isolated from sediment supply, and frommaritime influences, and thus vegetation becomesincreasingly terrestrial in nature and the duneincreasingly stable.

Progradation at the end of a spit may takeplace through longshore transport of sedimentand accumulation at the terminus of a sedimentcell and thus spits may extend and eventuallyenclose embayments. Successive recurves maybe preserved that mark previous positions of thespit terminus. Sediment abundance may lead toclosure of tidal inlets if a period exists in whichwave energy is able to seal an inlet against tidalcurrents. Such conditions are normally associatedwith storms (see below). If an inlet closes, aresidual body of sediment may be left in the tidaldeltas, which, in the exposed ebb delta, is likelyto be redistributed by wave action and retained

in the beach sediment budget. Tidal delta sandin the sheltered, back-barrier environment mayremain intact or be redistributed by locally gen-erated lagoonal waves (Cleary et al. 1976).

In areas where the modern coastal sedimentwas originally derived from a relict source (suchas the continental shelf, or reworked glacial sedi-ments), the supply may eventually be exhausted.Under such conditions coastal behaviour involvesreworking of a finite sediment volume, althoughit is inevitable that losses will occur throughleakage either offshore (to the shelf), alongshore(to adjacent systems) and/or onshore (to dunes).In the absence of sediment supply, a beach orbarrier may experience recession and landwarddunes then become new sediment sources as theyare eroded and their sediment is reintroducedinto the beach sediment budget.

As a finite sediment volume adjusts to variationsin wave energy along a beach, reorganization

WindSL

Mainland beach

Bay

Lago

on

Bedrock

Transgressivedune barrier

Progradedbarrier(Strandplain)

Stationarybarrier

Headland spit

Receded barrier

Mainland beach

Plan view Profile

Lagoonalsedimentoutcrop

Fig. 8.13 Features of temperate coasts associated with abundant and scarce sediment. Progradation isassociated with sediment retention in the beach andshoreface. With strong wind dispersal, a transgressivedune may form, dispersing sediment landward. Verticaldune growth may lead to a stationary barrier withonshore sediment transport. Longshore sediment supplymay enable spit elongation and formation of recurves as the spit terminus advances. With sediment scarcity,barriers migrate landward over underlying lagoonaldeposits. These may be exposed intertidally seaward ofthe barrier. Mainland-attached beaches may undergonarrowing and seaward dispersal of sediment. (After Roy et al. 1994.)

ESC08 9/7/06 3:20 PM Page 278


leads to localized erosion and accretion that isdependent on the ambient wave conditions andthe antecedent morphology. Such a coast mayexhibit spatially variable patterns of erosion andaccretion that are to some extent envisaged inthe coastal cell model of May & Tanner (1973).Long-term sediment scarcity may be manifest inbarriers retreating, such that back-barrier sedi-ments (e.g. lagoonal muds) are exposed on thebeachface (Fig. 8.13).

Migration of tidal inlets may take place undersediment abundance, which promotes longshoreextension of up-drift barriers, or under sedimentscarcity, as sediment is reworked to accommod-ate changing wave energy. The migration of inletscauses reworking of the associated tidal deltas(Reddering 1983). A suite of models of barrierplanform change has been presented (Carter et al. 1987; Carter & Orford 1993). Althoughbased on gravel barriers, a number of the changesenvisaged are evident on sand barriers, wherethe higher degree of dynamism makes changesless readily attributable to a particular forcingfactor. Cooper & Navas (2004) measured plan-form changes in a headland–embayment systemover more than 100 years, and attributed themto natural variations in bathymetry caused bysediment accumulation on the sea-bed. These, in turn, altered the pattern of waves approach-ing the shore, and changed the longshore driftdirections at the coast.

8.4.3 Response to storms and tsunami

Storms have been cited as the most importantsedimentary forces on temperate coasts. This is intuitively believable because of the high-magnitude processes that operate during stormsand the potential for dynamic thresholds to beexceeded that may dominate coastal behaviourat the historical time-scale. Their location out-side the tropics does expose temperate coasts tothe impact of cyclonic weather systems, as wellas to hurricanes that originate in the tropics butwhich make occasional landfall in temperatezones (Fig. 8.1). Storms produce large waves andelevated water levels (surge). Thus wave energyis not only increased but it operates at higher

levels than normal. Thus, sediments in dunesand back-beach areas are exposed to the effectsof waves and currents.

The importance of storms on boulder andgravel beaches has already been inferred, asthey produce conditions necessary for sedi-ment transport. In comparison, the importanceof storms on sand beaches is more difficult toisolate from the effects of other processes. Oneimportant consequence of storms on barriers is the occurrence of overwash, which moves sediment landward of the active beach profile,effectively removing it from the system until thebarrier migrates landward and it then re-entersthe system. Storms, which have been difficult to monitor because of the high energy levelsinvolved, have recently been identified as pro-ducing orders of magnitude differences in long-shore transport rates and even change in thenature of processes operating compared withlower energy conditions (Miller 1999). Storms,too, lead to enhanced energy levels in the surfzone. This has been associated with the enhancedlikelihood of infragravity waves and nearshorecirculations, which give rise to rhythmic shore-line topography (Komar 1998).

Storms have also been important in raisingwave energy levels such that they can overcometidal currents and close inlets. The Kosi Lagoonin South Africa closed as a result of a tropicalcyclone that temporarily enabled wave energyand wave-sediment transport processes to over-come the tidal currents generated by a massivetidal prism (Cooper et al. 1999). The ebb-deltawas thus mobilized and reworked landward to seal the former inlet. Without human inter-vention the system was unlikely to have everreopened. Storms have also been responsible forformation of new inlets in barrier island chainsthrough erosive overwashing, which has loweredbarriers to the extent that tidal flow is initiatedthrough low points (Fig. 8.14).

Large-volume, near instantaneous sand trans-port during storms has been responsible for majorchanges in beach erosion and accretion rates attidal inlets, with storms interpreted as the majordeterminants of tidal inlet behaviour at decadaltime-scales (Case Study 8.1). Similarly, Orford

ESC08 9/7/06 3:20 PM Page 279

280 ANDREW COOPER

Fig. 8.14 Inlet formed during Hurricane Isabelle, September 2003, Outer Banks North Carolina. The inlet was closed artificially within a few weeks in order to facilitate road access. (Photograph courtesy of the Program for the Study of Developed Shorelines, Duke University, NC.)

Case study 8.1 Impact of storms on sediment movement: Texas Coast, USA

Measurements of sediment movement along coasts can be made at various time-scales. Short-termstudies use instruments to provide information on the relationship between sediment transportand coastal dynamics. Longer-term studies involve comparison of beach profiles, maps, chartsand air photographs in order to quantify change, which then may be related to records of waveactivity, sea-level and climatic variables. In either approach, identifying the role of storms in thetotal sediment budget can be difficult because of their infrequency (which makes them difficultto capture on record) and their high magnitude (which destroys instruments).

Working on part of the barrier island coast of Texas (Case Fig. 8.1a), Morton et al. (1995)monitored beach sand volumes on two barrier islands separated by an inlet (Case Fig. 8.1b).The barrier islands are experiencing long-term sea-level rise and limited sediment supply andhence have been migrating landward over the past 100 years (Case Fig. 8.1b). The inner shelf ismuddy and all sand movements in the nearshore can be clearly identified and volumes of sandcalculated. Measurements over a 10 year period showed that volumes of sand lost from onebarrier island (Galveston Island) did not match the volume gained by the adjacent, down-driftbarrier island (Follets Island). Instead, sand eroded from the up-drift island was transportedacross the tidal inlet on the shoreface and deposited there. Subsequently it was transported bywaves onto the down-drift island.

ESC08 9/7/06 3:20 PM Page 280


Dis

tanc

e (m

)

−400

Date1930 1950 1970 1990

−300

−200

−100

0

100

200Follets Island

Car

la

Alic

ia

Dis

tanc

e (m

)

−300

Date1930 1950 1970 1990

−200

−100

0

200

300

(b)

Galveston Island

Adv

ance

Car

la

Alic

ia

100

Ret

reat

Adv

ance

Ret

reat

(a)

The study showed that sediment transfers were dominated by episodic storms that causednear-instantaneous erosion of barrier islands. A hurricane in 1983 (Hurricane Alicia) erodedbetween 50 and 70 m3 of sediment per metre of beach from Galveston Island and lesseramounts from Follets Island. Most of the eroded sand was transported south-west by strongcurrents associated with the hurricane and was deposited on the shoreface (only 12% of thevolume lost was carried over the islands and deposited as overwash).

Post-storm recovery (Case Fig. 8.1c) took place over several years and involved phases of sandmigration onshore, bar welding to the beach, and finally dune build-up as the beach became dry

Case Fig. 8.1 (a) Satellite Imageshowing barrier islands of the TexasCoast. Dominant longshore driftdirection is arrowed. Galveston Island and Follets Island are marked.(Image from NASA Earth ScienceApplications Directorate(https://zulu.ssc.nasa.gov/mrsid/).) (b) Long-term shoreline position(relative to 1930) of Follets Island andGalveston Island. Both islands haveretreated by over 200 m since 1930.(After Morton et al. 1995.)

ESC08 9/7/06 3:20 PM Page 281

282 ANDREW COOPER

et al. (1999) identified three periods of rapidforedune recession followed by long interven-ing periods of accretion, which they attributedto the impact of high-magnitude storms on a modally high-energy beach. The observationspointed to storms as the major determinants of historical scale coastal behaviour, as stormsproduced large-scale, near instantaneous ero-sional events. Interstorm periods were charac-terized by decadal-scale phases of post-stormbeach accretion.

Tsunami, as fast-moving solitary waves orgroups of waves, have been recorded on manytemperate coasts, but particularly on tectonicallyactive coasts, from sedimentary evidence (Dawson1994) and direct observation. These waves mayreach heights of up to 30 m as they approach theshore and cause overwashing of barriers andmainland shorelines. Typically tsunami depositscontain a range of grain sizes and may includemicrofossils that indicate a continental shelf origin (Hindson & Andrade 1999). They are

and wide. The beaches on Galveston Island never recovered their pre-storm volume whereas thoseon Follets Island continued to accrete as sand (carried south from Galveston during the storm)was transferred onshore under fair-weather wave conditions. The sand eroded from GalvestonIsland during the storm acted as a sediment supply for Follet Island and thus whereas Galvestonexperienced net retreat over the period 1980–1994, Follet experienced net progradation.

A subsequent hurricane in 1988 (Hurricane Gilbert) caused erosion and a temporary reversalin the recovery of Galveston Island, whereas on Follets Island, the post-Alicia sand still on theshoreface was pushed onshore, causing enhanced accretion. This study demonstrates the import-ance of antecedent conditions in determining coastal response to a storm. The accretion on FolletsIsland after Hurricane Alicia temporarily reversed the long-term erosional trend, however, whenthe sand supplied from Galveston Island is transferred onshore, it will be dispersed furthersouth in longshore drift and the long-term migration of the island is predicted to continue.

Relevant reading

Morton, R.A., Gibeaut, J.C. & Paine, J.G. (1995) Meso-scale transfer of sand during and after storms: implica-tions for prediction of shoreline movement. Marine Geology 126, 161–79.

(c)

−−−−

−−

Case Fig. 8.1 (c) Cumulativevolume changes of profiles measuredon Follets Island (closed circles) andGalveston island (closed triangles)relative to 1980. Note how FolletsIsland continued to accrete forseveral years after Hurricane Alicia in1983 whereas Galveston lost volumefor many years before gaining sand.(After Morton et al. 1995.)

ESC08 9/7/06 3:20 PM Page 282


less well sorted than deposits associated withstorms and may contain sequences depositedduring both landward and seaward return flows(Nanayama et al. 2000).

It should be clear from the above discussionthat the three-dimensional changes that occur in beach systems are multifarious in nature. Theprocesses that drive them (dynamics, sedimentsupply and antecedent morphology) are diversein type and in the time-scales at which they oper-ate. The very dynamism of these systems hindersattribution of cause and effect and, as such, theidentification of process–response relationshipson beaches and barriers is difficult. Such rela-tionships are perhaps most likely to be identifiedat the short time-scale through instrumentation.The difficulty lies in relating these observationsto longer time-scales because of the importanceof additional factors such as antecedent con-ditions and sediment supply and periodicity ofepisodic high-energy events. Relating site observa-tions to other locations is equally fraught withuncertainty because the particular constraintson sediment movement and geomorphologicalresponse are not conducive to full quantification.The up-scaling of short term observations tolonger time-scales suffers similar constraints forthe same reasons, coupled with the non-linearityof process–response mechanisms, unpredictability(or impossibility of complete measurement) offorcing mechanisms and morphodynamic feed-back. By its very nature, investigation of the roleof sea-level change and climate change requiresa long-term approach.


ACTIVITIES

A range of human activities influence the naturalfunctioning of temperate coastal systems. Per-haps the most pervasive impact results from the legacy of development at the shoreline. Thisimpedes the ability of the coastline to respondmorphologically to natural forcing because such responses may damage infrastructure. This situation also influences contemporary manage-ment approaches to the shoreline (see section 8.6).

In the pre-industrial era, coastal infrastructurethreatened by the sea was abandoned. Submergedformer landscapes and archaeological monumentsin the Middle East (Sivan et al. 2001) provideevidence of this practice. This still happens insome areas but the intensity of development andits human value (both social and economic) hasprompted a range of engineering interventionsto defend infrastructure.

The problems posed by human interventionarise partly from construction inside the activeprofile of the coastal zone (passive intervention)– for example, failure to recognize that dunesand/or the back-beach form an active part of the system during storms. Long-term stabilityduring interstorm periods may give a falseimpression of the coastal morphodynamic regime.Consequent removal of foredunes and theirreplacement with solid infrastructure rendersthat infrastructure vulnerable to erosion duringstorm conditions. Human alteration of sedimentmovement (active intervention) occurs throughstructures intended to alter wave and currentpatterns and intercept sediment in transport (e.g.groynes and jetties), and prevent sediment frombeing eroded (e.g. seawalls). Additional, lessobvious impacts on sediment movement includedenudation or stabilization of dunes. Such actionsalter the natural sediment budget.

Human activities may also impede the abilityof the shoreline to adjust to rising sea-levels.Developments built landward of the active coastalsedimentary system eventually find themselveswithin it, as a result of sea-level rise. Develop-ment at the shoreline imposes restrictions on theoptions available in response to rising sea-levels.This type of anthropogenic constraint on theboundaries of natural coastal systems poses amajor problem in coastal zone management andis often stated as the ‘erosion problem’. As notedabove, however, erosion is an entirely naturalpart of the cycling of sediment along coasts. Onlythe presence of human infrastructure renders ita ‘problem’.

There have been historical changes in humanuse of the coastline (Nordstrom 2000). The firstimpacts were probably vegetation destabiliza-tion in coastal dunes. Prehistoric occupation of

ESC08 9/7/06 3:20 PM Page 283

284 ANDREW COOPER

sand dunes was widespread and alternatingperiods of occupation and abandonment havebeen documented in many dune systems, whichprobably relate to alternating periods of stabil-ity and instability in which humans themselvesmay have played a role (Gilbertson et al. 1996).Wilson & Braley (1997) described the nine-teenth century engulfment of houses in Donegal,Ireland by blown sand as a result of overgrazingin adjacent dunes.

Navigation posed the next major human threatto natural coastal morphodynamics. Tidal inletsprovide strategic links between inland systemsand the open sea. Although inlets may be suffi-ciently deep, adjacent tidal deltas pose hazards to navigation. Dredging of channels and ebbdeltas was the initial response to this hazard.Such alteration of the sediment budget causesreadjustments, and, inevitably, the deltas reformand require further dredging. The next phase ofcontrol was the construction of jetties to fix thelocation of tidal inlets. In most cases, however,maintenance dredging is still required.

Increased levels of human utilization andoccupation of the coast have been accompaniedby developments in engineering that have pro-moted ever increasing interference with naturalcoastal functioning. Human interventions fallinto three main categories.1 Planned modification – where there is deliberateplanned construction of harbours, reclamationof coastal lands, sea wall construction, etc. Thisis almost always an engineered approach and is often planned around a relatively short-termtime-scale (decades). Generally, little attention isgiven to long-term morphodynamics at the site.2 Accidental modification – often a knock-oneffect from (1), where further along the coastthere is a direct impact on the general wave pat-terns and sediment transport pathways from,for example, an engineered structure. In manycases, simple ignorance of coastal processesleads to a direct modification of the coast. Forexample, the removal of sediment from beachescan have a beach-lowering effect and thereforean increase in beach vulnerability from storm-wave attack where less energy is dissipated andbackshore erosion takes place.

3 Reactive modification – in response to plannedor accidental modification. A process of progres-sive coastal modification can take place whenattempts to solve sedimentary problems usingengineering solutions produce further sediment-ary problems. These are then addressed by furtherengineering. Examples occur when longshoresupply of sediment is halted by stabilization oferoding cliffs that threaten human infrastructure.Beaches, deprived of sediment are then ‘stabilized’using groynes, which in turn starve other areasof the coast of sediment. Such situations produceprogressive down-drift intervention.

The above types of intervention take placethrough four main groups of human activit-ies, comprising coastal engineering, agriculturalactivity, the extractive industry and recreationalactivities. Each one of these will be addressed inturn but in many cases a combination of two ofmore of these activities may be present at thecoastal site and there may be feedback betweenthem as levels of activity change over time.

8.5.1 Engineering works

Off-site engineering has the potential to affectsediment supply to the coast. In cases wherefluvial sediment supply is important, construc-tion of impoundments, flow reduction and sediment abstraction from rivers can reduce thesupply and have an impact on the coastal sedi-ment system. At the Nile delta the reduction insediment supply (through impoundments) hasled to enhanced subsidence and relative sea-levelrise (Stanley & Warne 1998; see Chapter 7). Thesandy barrier islands at the leading edge of thedelta are subject to both impacts. In California,impoundment of numerous small rivers has ledto severe erosion on many beaches as the sandysediment supply has been reduced (Sherman et al. 2002). Another form of off-site impactmay occur through modification of back-barrierareas. Reclamation of the coastal fringe, forexample, can reduce the tidal prism and in turnmay induce changes in the tidal current regimeand the adjacent coast (Case Study 8.2). Coastalengineers typically attempt to modify coastalprocesses in order to protect property/structures

ESC08 9/7/06 3:20 PM Page 284


Case study 8.2 Shoreline changes near Wexford Harbour, Ireland

Wexford harbour in south-east Ireland (Case Fig. 8.2a) is a large estuary partly separated fromthe Irish Sea by a sand and gravel barrier. Associated with its wide inlet are extensive flood and ebb-tidal deltas. Locally generated south-easterly storm waves and diffracted ocean wavesproduce a northerly drift. The beach in the south of the bay is backed by cliffs of glacial till. A resistant bedrock headland (Greenore Point) in the south provides a hinge for development of a crenulate bay that extends northwards and forms Rosslare Bay. At its maximum extent, the spit that forms the northern part of this shoreline was 8.3 km long but lost about 3 km of itslength between 1925 and 1983 (Case Fig. 8.2b) as the spit was breached and eroded by storm-generated waves. Erosion of the spit was viewed with concern by the local authority because ofthe tourism associated with the beach, which was considered the only safe bathing area in theregion. A variety of small-scale coastal protection structures were emplaced (Case Fig. 8.2c).Groynes were emplaced to reduce longshore drift and rock armour was constructed adjacent tothe spit terminus in 1965–66. The apparent interruption of longshore drift by a pier at RosslareHarbour (constructed in 1882 and extended by 1902) was perceived as the reason for the erosion. A sandy beach built up to the south of the harbour. In 1978 the harbour was replacedby an impermeable breakwater and seawall.

Subsequent research (Orford 1988) suggests, however, that the shoreline erosion was re-lated to the realignment and failure of the spit rather than interruption of the longshore drift.Realignment of the coastal planform that included the mainland beach and the spit, by a

IrishSea

0 5 km

R. Sl ane

−10

N+

+

+

Study area

Rosslare Strand

Rosslare HarbourGreenore Point

CarnsorePoint

WexfordHarbour

(a)

TheRavenPoint

WexfordHarbour

RosslareSpit

RosslareBay

Wexford

−10

mde

pth

cont

our

1957

(b)

1950

1983

1925

0

N

0.5 1.0 km

?

H. W

.

Embankment

H. W.

Breach

Breach 1965–66

Stone groyne

Rock/tetrapod breakwater (1967)

Wooden groynes (1965)

RosslareSpit

Case Fig. 8.2 (a) Coastal setting of south-east Ireland showing Wexford Harbour, Rosslare Bay. (b) History of spitdisintegration: H.W., high water. (After Orford 1988.)

ESC08 9/7/06 3:20 PM Page 285

286 ANDREW COOPER

Case Fig. 8.2 (c) History anddistribution of coastal protection worksconstructed in response to coastalerosion. (After Orford 1988.) (d) Reclamation in Wexford harbourshowing initial intertidal area and post-reclamation area. The marked reductionin tidal prism was associated with majorchanges at the inlet and adjacent coast.(After Orford 1988.)

N

0 2Km

Rail link

1983

Protected

till cliff

Exposed

till cliff

Dune

Coast

RosslareStrand

Rock revetment

(c)

Dune blowoutPalisade fences

Woodengroynes

‘Private’ dunestabilization

Point source of beachnourishment

Wooden groynes(derelict)

Rip rap Sea

wall

RosslareHarbour

3.5 m yr−1

0.8 m yr−1

0.3–0.6 m yr−1

<1.4 m yr−1

<0.4 m yr−1

IntakeLand

R. Slaney

WEXFORD WEXFORD

Rosslare Harbour

N

Reclaimed

Reclaimed

Intertidal FlatsOld EmbankmentRailway Line

(i) 1845

(d)

(ii) 1958

60

180

60

Trainingwalls

18060

18060

60

60

ESC08 9/7/06 3:20 PM Page 286


or to improve navigation. There is often an asso-ciated undesirable sedimentary consequence.

8.5.1.1 Hard engineering

Groynes are shore-normal or oblique structuresthat intercept the longshore drift (Fig. 8.15). Theyare frequently constructed in areas deprived oftheir sediment supply by other human interven-tions. They can be successful in trapping sedimentwhere longshore sediment transport is naturallypresent (Komar 1998). At Cape May, New Jersey,construction of groynes to intercept longshoredrift led to recession of the down-drift coast byover 800 m (Pilkey & Dixon 1996). There havebeen attempts to lessen the impact of groynes ondown-drift coasts. For example, the vertical pro-files of some groynes have been adjusted as wellas their porosity in order to enable sedimentbypassing when the desired beach profile hasbeen achieved. A variety of different designs havebeen used, including changing the angle to theshore, and varying planforms (e.g. hammer-head

designs), all of which seek (with variable success)to lessen their undesirable effects.

Jetties also extend normal or oblique to theshore. Their purpose is to stabilize inlets and toprevent inlet migration. In so doing, they inter-rupt the longshore supply of sediment and causethe ebb-tidal delta to be destroyed and/or shiftposition. A range of ancillary activities usuallyaccompany jetty construction. These includechannel dredging and artificial bypassing of sediment around the inlet.

Sea walls are shore-parallel structures placedon a coast to prevent landward movement of theshoreline or to act as flood defences. Many sea-walls were built with the purpose of providingpromenades next to the sea. Various materials,slope profiles and slope angles are used to absorbwave energy. The sedimentary effects of seawalls(Fig. 8.16) include:1 beach lowering – where wave energy isreflected from the wall and sediment is strippedaway from the beach surface in front of thestructure;

few degrees, caused large-scale coastline retreat. Three sets of sedimentological informationsupported this conclusion. First, the spit failed by reduction in supply to its distal end that wasnot matched by reductions in supply at the proximal end. This was inconsistent with drift inter-ruption at Rosslare Harbour. Second, the relative immaturity of the bay planform is evidencedby observations of erosion and breaching before construction at Rosslare. The bay had notreached equilibrium with the ambient wave field. Third, a gross estimate of longshore trans-port supply from the eroding cliffs at Rosslare is approximately equal to the accumulation rates recorded south of the harbour and therefore there is not a major reduction in longshoresediment supply.

Orford (1988) also studied changes in nearshore bathymetry recorded on historic charts, andchanges in the estuary of Wexford Harbour. In the harbour, a large area of intertidal land hadbeen reclaimed in the mid-nineteenth century (Case Fig. 8.2d). The effect was to reduce the tidalprism of the estuary. This, in turn altered the balance of tidal and wave power at the inlet andcaused large-scale sediment reorganization as the positions and strengths of the main flood- andebb-directed currents shifted. These ultimately caused the demise of the spit and the changes inplanform at Rosslare.

Relevant reading

Orford, J.D. (1988) Alternative interpretations of man-induced shoreline changes in Rosslare Bay, southeastIreland. Transactions of the Institute of British Geographers 13, 65–78.

ESC08 9/7/06 3:20 PM Page 287

288 ANDREW COOPER

2 edge erosion – where localized turbulence atthe end of the wall causes severe erosion;3 potential grain-size changes – natural sedi-ment sorting can be significantly altered by theemplacement of a sea wall;4 separation of dune and beach systems – wherea seawall removes any dune or back-beachmaterial from the active profile and precludesincorporation of such sediment into the profileduring storms – this also causes loss of sedimentsupply from the beach to adjacent dunes;5 new littoral currents and sediment transportmay be generated by the presence of sea walls,which in turn create additional problems forsites further along the coast, and may threatenthe stability of the seawall itself.

Seawalls are commonly used to protect infra-structure from long-term coastal retreat. In manycases they are also used inappropriately to com-bat seasonal erosion. More often than not thisinduces irreversible changes on the coast. Perhapsthe greatest impact of seawalls is in fixing thelandward boundary of the coastal sediment sys-tem. Under a rising sea-level (see 8.7) landwardmigration of beaches is precluded. Thus beaches

backed by seawalls become narrower and steeper,often disappearing altogether.

Offshore breakwaters are designed principallyto modify wave patterns and to produce sedi-mentary effects on the shoreline. They absorbwave energy before waves reach the shore. This effect reduces onshore–offshore sedimenttransport (especially during storms) but can, incertain circumstances, reduce littoral drift andcreate cuspate accumulations or tombolos as a result of wave refraction effects. Attempts tosimulate natural log-spiral forms of headland–embayment cells using artificial headlands have been made using both attached and off-shore headlands (Silvester 1976). Tombolos and salients have also been produced throughinstallation of offshore structures (French 2001).As with all engineering activities in the coastalzone, unexpected coastal events can cause undesired morphodynamic effects. Under stormconditions, for example, the water layer abovethe crest will no longer intercept energy fromincoming waves. The height of the breakwaterposes serious problems because if too low, in-adequate protection will ensue, and if too high,

Direction of natural beach sand transport

Groyne

(a)

(b)

(c)

Duneline

Groynes trap sand,erosion accelerated

Formershoreline

Remnants of early groyne

Formershoreline

Duneline

Duneline

Progressive need for additional groynes

Fig. 8.15 Typical impacts and temporal pattern ofgroyne construction. (After Pilkey et al. 1998.)

ESC08 9/7/06 3:20 PM Page 288


interference with the shore processes will resultin new problems.

8.5.1.2 Soft engineering

Beach nourishment is a relatively recent tech-nique of shore protection and following someearly twentieth century emplacements (Bird 1996)widespread application began after the ‘AshWednesday Storm’ of 1962 on the eastern USA.Beach nourishment tackles erosion by replacinglost sediment and relies in part on natural waveaction dispersing introduced material along theshoreline and achieving equilibrium with ambi-ent dynamics. Ideally therefore, the introducedsand should replicate the native sand texture

(Dean 1974). Beach nourishment often goes handin hand with navigational channel maintenancewhere dredge spoil provides suitable sediment.A variety of emplacement modes have been employed to nourish beaches. They include: single direct emplacement from offshore dredgers,followed by mechanical profiling; trickle feed-ing, where sediment is introduced from a singlesource at small volumes; recycling of sedimentfrom down-drift sinks to up-drift source areas;and emplacement in the nearshore for naturalonshore transport (French 2001).

Nourishment has been seen as a panacea for erosion control. Once emplaced, nourishedbeaches adjust to natural dynamics and mor-phology by dispersing sediment and producing

Dune scarp

Beach narrowing

Beach gone

High-rise development

1. Before the wall

2. Wall constructed.Development proceeds as buyers believeproperty protected by the wall

3. Two to forty years later

4. Ten to sixty years later

Steepening ofoffshore slope

Offshore slope has steepened

Original oceanfront house destroyed in large storm. Bigger, 'better' seawall built

As slope increases,wave size increases.A higher wall is needed

Fig. 8.16 Impacts of seawalls. Seawalls fix the landwardmargin of the beach, thereby inhibiting its ability to respondto storms or sea-level rise. They also may be constructed onthe beach, thus causing effective loss of the active beach.Active loss occurs as waves reflect off the seawall and erodethe beach, particularly during storms. Over the mediumterm, seawalls cause beaches to narrow as waves arereflected and sediment is lost. The beach may disappear andthe seawall itself may be undermined as the shorefacesteepens. (After Pilkey et al. 1998.)

ESC08 9/7/06 3:20 PM Page 289

290 ANDREW COOPER

geomorphological forms (planform and profile)in response to wave energy. If sediment is re-moved from nearshore areas, the net effect is to steepen the nearshore profile, which in turnalters the relationship between beach morpho-logy and wave dynamics. In certain cases, off-shore breakwaters are constructed in front of thenewly nourished beach to help break up incom-ing wave energy in order to reduce erosionaleffects. Scarcity and cost of available sediment is a major constraint on beach nourishment.There is an inherent lack of accuracy in pre-dicting the longevity of nourished beaches and thus their cost–benefit relationship can bedifficult to assess (Pilkey & Dixon 1996). Beachnourishment has the effect of widening the drypart of the beach, which in turn may lead toenhanced aeolian activity. Engineering historyhas shown that there is a strong regional differ-ence in the durability of nourished beaches alongthe eastern coast of the USA. Leonard et al.(1989) noted that beach durability was prob-ably related to many factors, most of whichcould not be isolated, but that storm frequencyand overall wave energy were probably particu-larly important.

Other forms of soft engineering involve the use of ecological elements to deliberatelyinduce morphological change. Examples includevegetation planting on foredune systems toinduce additional sediment build-up (Woodhouse1978) and to provide additional buffering againststorm attack. Similar effects can be achieved byconstruction of various forms of wind trap thatencourage aeolian sediment to accumulate. InIreland, non-indigenous species (sea buckthorn,laurel and sycamore) have been planted in sanddune systems in order to stabilize blowouts and block access. In certain cases a scenario of‘overmanagement’ may develop in coastal sanddunes. Attempts to stabilize the dune morpho-logy have sealed the main landward corridors of sediment transport. The US National ParksService artificially built up many foredune sys-tems through trapping and marram planting onbarrier islands during the 1930s. This resultedin a deficit in natural sediment and nutrientthroughput within the dunes. In turn, this induced

a massive reduction in washover and blow-overprocesses and increased beach erosion on theseaward margin. By the 1970s there was a majorreappraisal and a decision was made to allowthe barriers to readjust naturally.

8.5.2 Agricultural activities

Domestic animals (sheep, cattle, horses, etc.) are frequently grazed on coastal dunes. Theimpact on the dune surface depends on stockingdensities. Trampling and grazing can result invegetation damage. The resulting instability maycause large-scale wind blow in dune systems. Lowstocking levels can promote vegetation stabilityby stimulating vigorous growth via grazing andthe addition of manure. Rabbits were importedfor food into coastal dunes in Ireland from the twelfth century onward. Rabbit burrowingsubsequently led to the destabilization of sanddune systems and the dynamics of the rabbitpopulation became a major factor in dune systemdynamics. The myxomatosis-induced popula-tion crash in the 1950s led to enhanced dunestability. The presence of bare sand in dunes hasoften been seen as an opportunity for large-scaleafforestation. In Spain and Portugal significantstretches of coast have been planted with pinetrees, which have resulted in almost total block-age of inland sediment movement.

8.5.3 Anthropogenic influence on sediment supply

Human influences on sediment supply can besubstantial. The manner in which humans altersediment supply and transport at the coastincludes the following:1 reduction or increase in sediment supplythrough river regulation and land-use change;2 loss of sand by dune destruction or stabilization;3 loss of sediment through dredging (navigationor aggregate extraction);4 blocking of littoral transport by structures;5 loss of sediment by stabilization of cliff sources.In many cases, agricultural practices alter sediment supply. In central Italy widespreadcoastal erosion post-1950 was ascribed to gravel

ESC08 9/7/06 3:20 PM Page 290


extraction and land surface stabilization underagriculture (Coltorti 1997). Similarly, dams onrivers have been responsible for reduction ofsediment supply through direct entrapment andthrough reduction of transport capacity by redu-cing water flows (Case Study 8.3), and large-scaleurbanization may reduce sediment dischargefrom paved areas. Mining and quarrying may alsolocally alter sediment supply. The disposal ofmine waste in Corsica introduced vast quantitiesof sediment to the beach systems and caused

shoreline accretion of over 400 m (Bernier et al.1996). More widespread sediment inputs to thecoastal zone arise from waste disposal – fewbeaches do not contain occasional bricks, glassfragments or other human debris.

Small-scale sediment removal from beaches anddunes has taken place for a variety of reasons,including aggregate, construction, fertilizer andanimal bedding. Carter et al. (1991) reported5000–6000 t of sand removal over a 10-yearperiod from eleven Northern Ireland beaches and

Case Study 8.3 Impacts of reduced fluvial sediment supply to California beaches, USA

The sandy beaches of southern California are an important economic resource. Over the pastfew decades many beaches have experienced severe erosion that has reduced their amenityvalue and placed human infrastructure at risk. Griggs & Savoy (1985) estimated that > 80% ofthe California coast was eroding and c. 30% was in the high risk category. The reasons for therapid coastal erosion relate largely to reduction in fluvial sediment supply, which was estimatedto yield 70–85% of all beach sand (Sherman et al. 2002) (Case Fig. 8.3a). Additional reductionsin sediment supply relate to interruptions of alongshore sediment movement by engineeringstructures (mainly jetties) and protection of eroding bluffs by seawalls (Liedersdorf et al. 1994).The fluvial sediment supply reduction was related to the construction of dams on sediment-yieldingrivers in the steep, arid California hinterland (Case Fig. 8.3b). More than 500 dams impoundmore than 42,000 km2 (38% of the surface area) of California (Willis & Griggs 2003). Shermanet al. (2002) calculated the sediment retention in 28 dams and 150 debris basins in southernCalifornia and concluded that these structures impounded > 4 million m3 of sediment per year,equivalent to 3 m3 of sediment per metre of shoreline in the five southern coastal counties in the State. Willis & Griggs (2003) determined that 70 dams (13%) were responsible for 90% ofsediment reductions to the coast (Case Fig. 8.3c). Half the dams are old and have lost significantwater retention capacity as a result of sedimentation and are also in need of maintenance. Inview of their reduced functionality, a number of dams have been identified for removal.

In the face of rapid beach erosion, a study on the management options was undertaken(Coyne & Sterrett 2002). This study identified an annual tax revenue to the State of $4.6 billionbased on beach tourism and recreation. Potential losses in revenue through beach erosion werecalculated at $1 billion in taxes. There was thus a strong economic argument for beach restora-tion. The study recommended a twofold approach to beach restoration involving opportunisticbeach nourishment and dam removal. Opportunistic beach nourishment involves the emplace-ment of ‘sand of opportunity’ that becomes available from construction or excavation. Previousexperience had shown such emplacements to be successful in improving beach longevity. Themore far-reaching recommendation is the removal of dams that are no longer serving a usefulfunction, in order to increase the natural sediment supply to beaches. The economics of beachesversus the reduced value of dams was probably an important factor in enabling this manage-ment strategy to be adopted. A number of studies are presently underway in advance of damremoval on several sediment yielding rivers.

ESC08 9/7/06 3:20 PM Page 291

292 ANDREW COOPER

Case Fig. 8.3 (a)Sediment cells on thesouthern California coast.Each cell has identifiablesediment sources, sinksand transport pathways.(Adapted from Flick1993.) (b) Rivercatchments and dams inthe Los Angeles sector ofthe San Pedro cell (see(a)). The large number ofdams in the small, steepcoastal catchments hassignificantly reducedsediment supply to thecoast and contributed tocoastal erosion. (c) Littoralsediment budget forCalifornian beaches. Ifsediment flux QL in (i)(natural conditions)diminishes to qL in (ii)(post-impoundment) andlittoral transport QLT1remains constant theinitial beach volume mustreduce to balance thesediment budget. (AfterWillis & Griggs 2003.)

N

Los Angeles

0 50 km

RiverSandy beaches

Submarine Canyon

Submarine basin

Littoral cell

Rockycoast

Pt.Conception

San Miguel I.

Santa Rosa I.

Santa Cruz I.

Santa Barbara

Santa Barbara cell

Santa Monica cell

San Pedro cell

Oceansidecell

Santa Monica Basin

SanPedro Basin

Santa Catalina I.

San Diego Trough

CatalinaBasin

San Clements I.

Los Angeles

Riverside

San Gabriel

N 0 10 km

Puddingstone

Hansen

Sepulveda Santa Fe

Prado

Pacific

(b)

(a)

Ocean

Big Tujunga

Devil's Gate

(i)

(c)

OceanOcean

(ii)

t= + −

Land

BeachVB

QL

VB

QLT1 QLT2

q L

QLT1 QLT2

VB QLT1 QL QLT2t

= + −VB QLT1 qL QLT2

ESC08 9/7/06 3:20 PM Page 292


Brunsden & Moore (1999) noted large-scalegravel removal from Chesil Beach. In areas offinite sediment volume, this may produce a seri-ous erosion problem. Large scale, commercialbeach and sand dune mining is also practised. In Australia and South Africa, natural concen-trations of heavy minerals in dune sands areexploited. Offshore aggregate removal is a com-mercial operation that has led to a reduction in the nearshore sediment supply and also tochanges in bathymetry (see Chapter 10). Thesein turn alter wave refraction patterns and maycause readjustments in shoreline morphology.The English village of Hallsands was eroded as a result of offshore aggregate extraction for harbour construction (Pearce 1996). Nearshoresediment extraction also takes place in the searchfor precious minerals (e.g. diamonds off theSouth African and Namibian coasts).

8.5.4 Recreational activity (human trampling,leisure vehicle activity)

The coastal zone has traditionally attractedhuman visitors who have utilized it as a foodsource, living area and more recently as a leisurearea. Recreational activities can be a major fac-tor in coastal sedimentology. In coastal dunes,human trampling and use of recreational vehiclescan result in large-scale reactivation of fixedsediment (Gilbertson 1981). Car parking onbeaches can induce compaction of the sedimentand interfere with natural sediment transport in the intertidal and adjacent supratidal zones.

‘Cleaning’ beaches of litter including seaweedon drift lines can cause loss of potential dunenucleation sites and nutrients needed to pro-mote plant growth.


8.6.1 Management approaches

Management of sedimentary coastlines involvesconsideration of (i) utilization of the coast suchas to minimize human impact on the sedimentarysystem, and (ii) balancing human utilization witha naturally dynamic system. Management of sedi-mentary coastlines must therefore begin with anunderstanding of the processes that shape andmaintain those coasts at time-scales relevant tomanagement. The discussion above should imme-diately make clear that this is a difficult task. Therange of processes and the physical constraintsthat operate at varying spatial and temporalscales are unlikely to be well understood at anycoastal location, and hence a wide margin oferror is inherent in sedimentological assessmentsrelated to coastal management. Existing con-straints on a coastline imposed by human inter-ventions (Fig. 8.17a), and the potential impactof further intervention on adjacent stretches ofcoastline, impose additional considerations formanagement of sedimentary coasts.

Applied sedimentological investigations inthe coastal zone are therefore typically under-taken for two purposes. One is the design and

Relevant reading

Coyne, M.A. & Sterrett, E.H. (2002) California Beach Restoration Study. California Department of Boating andWaterways and State Coastal Conservancy, Sacramento, CA. (Available at: http://dbw.ca.gov/beachreport.htm)

Griggs, G.B. & Savoy, L. (1985) Living with the California Coast. Duke University Press, Durham, NC.Flick, R.E. (1993) The myth and reality of southern California beaches. Shore and Beach 61, 3–13.Liedersdorf, C.B., Hollar, R.C. & Woodell, G. (1994) Human intervention with the beaches of Santa Monica

Bay, California. Shore and Beach 62, 29–38.Sherman, D.J., Barron, K.M. & Ellis, J.T. (2002) Retention of beach sands by dams and debris basins in

southern California. Journal of Coastal Research (Special Issue) 36, 662–74.Willis, C.M. & Griggs, G.B. (2003) Reductions in fluvial sediment discharge by coastal dams in California and

implications for beach sustainability. Journal of Geology 111, 167–82.

ESC08 9/7/06 3:20 PM Page 293

294 ANDREW COOPER

(a)

(b)

(c)

Fig. 8.17 (a) High-rise tourist developmenton a barrier spit at Mar Menor, south-eastSpain fronted by a narrow, eroding beach.Such intensive development inhibits theability of the shoreline to respond to changesin sediment volume. (b) Coastal defencestructures on a shingle beach in Kent,England. A series of stone and woodengroynes has been used to trap and retainshingle and a revetment has beenconstructed on the back-beach, backed by an armoured slope. This ‘hold the line’option seeks to counteract natural shorelinebehaviour in the face of reduced sedimentsupply and relative sea-level rise. (c) Rapidcoastal retreat on the Outer Banks of NorthCarolina (USA) near Nags Head has erodedthe sand dunes on which these houses wereconstructed. The concrete septic tank wasoriginally buried in dunes. The constructionof sea walls is prohibited in this State andthus the coastline can migrate at the expenseof poorly located developments. The housesare now abandoned and must be removed bythe owners. This approach is consistent withthe ‘managed retreat’ or ‘do nothing’ coastalmanagement option.

ESC08 9/7/06 3:20 PM Page 294


environmental assessment of specific engineer-ing projects, and the other is the development of management policies in response to environ-mental change. These are outlined briefly belowand the techniques in use are described in thefollowing section. The design and impact assess-ment of coastal infrastructure involves predictionof coastal processes and sedimentary responsein the face of anthropogenic activities. Typicalconsiderations include the design of groynes to interrupt longshore drift and the likely sedi-mentary consequences of their installation atrelevant time-scales. In the face of actual orpotential coastal change four policy options arecommonly identified. These are as follows:1 do nothing;2 hold the existing line;3 advance the line;4 retreat.

The decision-making process related to coastalmanagement policy of this type varies in itsdegree of formality and, depending on the localresources and infrastructure at risk, the coast-line will be defended or otherwise. The approachhas been formalized in the concept of shorelinemanagement plans (SMPs) in England and Wales,whereby each section of the coast is subject to a range of analyses designed to inform policychoice. The ‘do nothing’ policy option is one inwhich natural processes are left to operate freeof further human intervention. This option istypically adopted where no coastal infrastruc-ture is at risk and permits the coast to fluctuatefreely within current constraints.

In the ‘hold the line’ option, the loss of infra-structure is considered to be unacceptable andthe coast must be defended. The Netherlandsdecision to maintain the 1991 shoreline positionis an example of such policy at a national level(De Ruig & Hillen 1997). This option is usuallyassociated with solid structures, although thebeach nourishment approach is becoming morewidespread. It is an unfortunate but unsurpris-ing fact that most landowners prefer the ‘defend’option, particularly if the expense is borne byothers (Fig. 8.17b).

The ‘advance’ option is one in which a deliberate decision is made to claim intertidal

or subtidal land and to defend this advancedposition. In practice this option is rarely takenon the open coast and is confined mainly to areasthat are actively prograding, and where infra-structural development follows the advancingshoreline. An important consideration in thispolicy is whether the progradation is likely to be sustained.

The retreat option is one in which the inevit-ability of shoreline retreat is identified andaccepted. Structures that are undermined or collapse are not replaced, and might even bedeliberately removed, and human infrastruc-ture is relocated landward (Fig. 8.17c). Statutesthat prevent seawall construction are usuallyindicative of the adoption of a ‘retreat’ strategyas in Maine or North Carolina (Pilkey et al.1998). Selection of the retreat option is typic-ally taken on economic grounds although it isincreasingly popular owing to the conservationbenefits of natural coastal systems. It is polit-ically the most difficult option to pursue wheninfrastructure is at risk.

8.6.2 Sedimentology in coastal zone management

Understanding the sedimentary dynamics of thecoastline is central to the adoption of manage-ment policy and design of human infrastructure.For the purposes of coastal management, thetime-scale of interest is typically in the range of years to decades. A range of approaches areavailable to the applied sedimentologist to developan understanding of the morphodynamics of tem-perate coastlines, and more than one approachmay be necessary. The inability to upscale short-term measurements to longer time-scales is a keyconstraint on modern applied sedimentology.This in part lies in the feedback relationships andcomplexity of processes in the nearshore zone. It should therefore be acknowledged at the outset that it is impossible to quantify coastalbehaviour in a generic sense and that coastalmorphodynamics at meaningful time-scales willalways be expressed in qualitative terms. Withthis caveat, three types of applied sedimento-logical study typically inform coastal manage-ment: field studies, historical geomorphological

ESC08 9/7/06 3:20 PM Page 295

296 ANDREW COOPER

change/geoindicators and modelling. A key con-sideration of such studies are the relationshipsbetween sedimentary processes and geomorpho-logical change at different time-scales.

8.6.2.1 Field studies

Field studies in applied coastal sedimentologyutilize a wide range of techniques including, atthe short time-scale, tracer studies of grain move-ments, sediment traps that capture some or all ofthe sediment flux at a given point and acousticand optical backscattering to measure sedimentconcentrations. Each of these measures is usu-ally accompanied by quantification of some ele-ment of the dynamic environment (e.g. waveparameters, currents) or geomorphological re-sponse (e.g. bed-level changes), and relationshipsare sought between dynamics and sedimenttransport. Over longer time-scales, repeat fieldmeasurement of beach profiles, nearshore topo-graphy or other elements of the coastal morpho-logy are used to assess sedimentary behaviourover time. These too are often accompanied bymeasures of dynamic data or proxy dynamicdata, usually at longer time-scales than thoseutilized in short-term studies. Recognition of the lack of a direct relationship between shortand medium-term data hampers the integrationof data gathered using these approaches (Carter& Woodroffe 1994).

As field data reflect a range of local variables(textural, inherited factors and dynamics), theirgeneric applicability remains to be tested (Cooper& Pilkey 2004a). A range of analytical tech-niques (visual, statistical, mathematical) are usedto test for relationships between sedimentary/morphological factors and dynamics (e.g. Clarke& Eliot 1988). Unfortunately, the results of fieldstudies are often viewed as universal and empir-ical relationships are applied elsewhere withoutdue consideration of their limitations (Cooper& Pilkey 2004b). Faced with the difficulty ofrelating empirical measurements of coastal changeto measurements of forcing factors, ever moreelaborate approaches have been developed. Theseinclude neural networks (Chen et al. 1990), whichseek patterns in non-linear systems including

observed sequences of coastal morphologicalmeasurement. Such approaches are often con-strained by lack of sufficiently long-term datasets to render the approach statistically valid.Constraints on these approaches relate to therepresentativeness of the data collected and itsspatial coverage. Most importantly, the effect ofstorms is often missed.

8.6.2.2 Historical change and geoindicators

A range of indicators of coastal morphologicalchange is available for most coastlines. Theseare variable in quality and temporal/spatial coverage. Using long data sets of beach changeand storminess, Bryant (1988) ascribed changesin high-tide morphological state to a combina-tion of rainfall, storminess, air circulation andsea-level change. The Dutch beach nourishmentapproach (Verhagen 1992) involves initially obtaining and analysing a decadal record of weeklybeach profiles on the beach to be nourished. Theassumption is made that the nourished beachwill behave more or less like the natural beach.In the field, a range of indicators exist includingsediment accumulation against groynes, rates ofinlet movement, etc. (Bush et al. 1996). On manybeaches, the presence of groynes, jetties, fishingpiers, erosion debris, seawalls and other engineer-ing projects provides geoindicators. Constraintson the historical approach relate to the availabil-ity of data, the extent to which coastal behaviouris captured by a series of snapshots, and whetheraccompanying dynamic data exist to aid inter-pretation of the morphological information.

8.6.2.3 Modelling

Modelling of coastal processes is a widespreadapproach in contemporary coastal management.Models fall into two categories: (i) researchmodels and (ii) applied or practical models(Davies & Villaret 2002). Differences betweenthe two are discussed by Thieler et al. (2000).Research models serve the purpose of investigat-ing the mechanisms of sediment transport in the coastal zone and can yield important newdiscoveries regarding the role of interacting waves

ESC08 9/7/06 3:20 PM Page 296


and currents, bed friction and sediment textureon, for example, nearshore sediment transport.Applied models, in contrast, set out to deliver a quantitative prediction of nearshore sedimenttransport or geomorphological response. Appliedmodels are deployed in a routine manner thattypically involves testing against a set of field data.Differences between the predicted and observedchanges are then considered and coefficients areapplied to render the two as close as possible forthe historical data. This produces a ‘calibrated’or ‘tuned’ model that is then applied to design astructure, to determine its environmental impacts,or to predict future shoreline evolution.

There are several types of coastal appliedmodel in use. The most commonly used processmodels in the USA for example are SBEACHand GENESIS. Their formulation is public andtherefore can be subjected to public scrutiny(and found wanting – Thieler et al. 2000). InEurope, many of the applied coastal models areproprietary software, developed and used in acompetitive environment by coastal engineer-ing firms. The models are often sold and used byother consultants. In this environment, wherethe precise formulation of the models is not public, such models (e.g. LITPACK, UNIBEST)operate as black boxes. In nature the variabilitybetween beaches and even within a single beachis so great that the parameterizations on whichapplied models are based are not transferablebetween sites. This seldom contemplated fact is somewhat masked by the calibration process,which in fact involves the use of a constant(‘fudge factor’) to achieve agreement betweenpredicted and actual change. The ‘calibrated’model is then used to predict future scenarios.

Modelling is characterized by endless attemptsto compare model output with field data (Mulderet al. 2001). This emphasizes on the one handthe paucity of good field data sets, and on theother the wide variety of model types. In spite ofefforts to test models against field data, modelrefinement is fundamentally prevented by theinability to resolve the spatial variability in factors that control sediment transport and the natural temporal variability in these factors.Models themselves thus do not provide any

more reliable predictions of coastal behaviourthan other measures. This fact is not commonlyappreciated. Even in the comparatively simplecase of aeolian sediment transport, aeolian trans-port modelling has had little success in predictingactual transport rates at even short time-scales(Sherman et al. 1998).

Modelling approaches are also used in longer-term coastal evolution studies. Here, in spite ofa lack of scientific validation, the Bruun ‘Rule’model (see 8.7) is the most widely appliedapproach in predicting coastal response to sea-level rise. The shoreline translation model, atwo-dimensional model of coastal profile change(Cowell et al. 1995), utilizes the Bruun approachin maintaining a consistent nearshore profile asthe coastline responds to sea-level rise.

8.6.2.4 Composite approaches

Understanding the sedimentology of temperatecoasts is an imprecise science that is informed bya range of potential data. Each dataset providesonly a piece of the necessary information forcorrect interpretation and it is important thatthis is considered in applied sedimentology. Thefullest understanding of coastal morphodynamicbehaviour is therefore likely to be achieved bythe compilation of data from a range of sources,which when compared against each other forconsistency or opposing trends, enable descrip-tion of coastal behaviour.

Understanding coastal sedimentology at time-scales useful to humankind requires utilization ofa range of data of varying quality and coveringvariable time periods. A composite approachusing historical, model and field observationscould provide the most comprehensive assess-ment of coastal sedimentary behaviour. In such studies ‘order of magnitude’ or qualitativeanswers are sought that aid understanding ofthe processes operative at varying time-scales.

Building on this type of composite approach,a novel initiative in future shoreline predictionhas been undertaken in Great Britain. Termed‘Futurecoast’ (Cooper & Jay 2002) this approachuses a combination of geomorphological ex-pert opinion (delimitation of cell boundaries),

ESC08 9/7/06 3:20 PM Page 297

298 ANDREW COOPER

historical change records, wave modelling, morphological measurements (e.g. beach width) and bathymetric changes (sediment budget guide)to assess the likely future behaviour of the coast. Future coastal evolution is assessed fortwo scenarios: unconstrained (i.e. assuming nodefences or new management practices) andmanaged (i.e. assuming present managementpractices continue indefinitely). The objective isto describe the ‘behaviour’ of a beach system inits coastal context and thus describe the likelyfuture ‘behaviour’ of the same beach.

8.7 FUTURE ISSUES

Two major issues dominate future concernsabout temperate coastlines – climate change and sea-level change. In both instances, studiesof coastal response to past changes provideinformation on likely future scenarios. As thehuman population is now greater than ever, andis increasingly concentrated in the coastal zone,options for dealing with future changes in theshoreline must consider this constraint.

The sedimentary effects of long-term climaticfluctuation have been analysed in several temper-ate coastal systems. Two distinct climatic phaseshave been noted globally in the late Holocene –the Holocene Climatic Optimum and the LittleIce Age. The effects of both have been identifiedin coastal dune sequences, where climatic deter-ioration (cooler and with increased storminess)is associated with large-scale instability anddevelopment of transgressive sand sheets, andclimatic amelioration is linked to enhanced vegetation growth and hence stability (Gilbertsonet al. 1996; Wilson et al. 2001). At the decadaltime-scale, variations in climate are character-ized by ENSO (El Niño–Southern Oscillation)and the NAO (North Atlantic Oscillation). Bothoscillations have been associated with changesin coastal behaviour at such time-scales. Goy et al. (2003) ascribed emplacement of beachridges in southern Spain over a 400-year periodto fluctuations in the NAO, with emplacementduring stormy periods separated by periods ofshoreline stability in intervening calm periods.

Sea-level change is a key issue for temperatecoastlines, and is widely associated with problemsof coastal erosion. A global predominance ofrising sea-levels means that this area has receivedmost attention, however, many temperate coastsare experiencing sea-level fall. As sea-level simplymediates the level at which short-term dynamicprocesses operate, distinguishing the sea-level-related signature of coastal change from otherforcing mechanisms discussed above (see sec-tion 8.4) is not straightforward. Nonetheless anumber of models of coastal response have beenproposed based on field observations, analysesof historical change and laboratory studies. Anadditional source of information lies in strati-graphical studies (typically spanning severalmillennia). A number of idealized modes ofresponse to sea-level change have been identified.For a rise in sea-level, simple two-dimensionalshore-normal models (Fig. 8.18) involve eitheran erosional response, a rollover response, or in situ drowning (overstepping).

Sea-level 2Sea-level 1



(a) Erosional

(b) Rollover

(c) Overstepping

Wave base 2Wave base 1

ErosionVolume eroded

Volume deposited

Onshore transport distance

Fig. 8.18 Erosional, rollover and overstepping models ofresponse to sea-level rise. The erosional response (a) involvesseaward dispersal of sediment to raise the nearshore profile intandem with sea-level rise. The coastal migration model (b)envisages retention of a fixed sediment volume that migrateslandward via barrier overwash. With no return mechanism, the barrier ‘rolls’ landward to maintain a fixed volume. In theoverstepping mode (c), barriers are unable to respond sufficientlyrapidly to sea-level rise and are drowned. (After Carter 1988.)

ESC08 9/7/06 3:20 PM Page 298


The erosional response model of sea-level rise is one that envisages maintenance of thenearshore profile by shoreface accretion at theexpense of beachface and berm/dune erosion asthe beach profile moves upward and landward.This simple model envisages a layer of sedimentbeing preserved on the shoreface as sea-levelchanges. A numerical form of this response wasoptimistically termed the Bruun ‘Rule’ (after theengineer who invented it) by Schwartz (1967),although subsequent studies have shown that, in addition to the lack of field support, it is aflawed concept because of the very restrictedconditions in which it is likely to operate and the limited range of processes it considers (Listet al. 1997; Pilkey & Cooper 2004).

The rollover response takes place on low barriers where waves overtop the berm, therebyeroding and transporting beachface sediment to the back-barrier area. In a sediment-limitedenvironment this causes the barrier to migratelandward. The migration rates of such systemsare dependent largely on the slope of the under-lying surface over which they migrate. In situdrowning is a phenomenon of barriers thatrespond more slowly to dynamic forcing thanthe rate of sea-level rise. Thus the system doesnot reach equilibrium with a change in sea-levelbefore it is stranded below the level of effec-tive coastal processes. Examples cited includegravel barriers, which have a response time thatis slower than sand barriers and are thus leftstranded on the shelf as sea-level rises (Carter & Orford 1993). Diagenesis (beachrock andaeolianite formation) may fulfil a similar func-tion and several drowned shorelines on the SouthAfrican and Australian coasts are preserved asbeachrock/aeolianite ridges (Cooper 1991).

Studies of stratigraphical successions that haveresulted from sea-level rise have identified dif-ferent modes of preservation depending on therate of sea-level rise and the supply of sedi-ment (Belknap et al. 2002; Thom 1984). At themillennial scale, the balance between the rate of sea-level rise and sediment supply are thedominant controls on coastal evolution. At thesetime-scales, variations in wave dynamics areconsidered to be masked by the other processes

(interestingly, the reverse is true of shorter termobservations).

The response of coastal dunes to changingsea-level has received comparatively little atten-tion other than their consideration as part of the beach profile. Dunes are, however, subjectto a range of distinctive processes that may varyas sea-level changes. Carter (1991) presented aset of models of potential dune response to sea-level change that incorporated consideration of sediment supply and vegetation cover, twofactors that are critical in dune sedimentology(Fig. 8.19).

Changes of beach planform in response tosea-level changes have been much less studied,and are more difficult to relate to sea-level changesas opposed to temporal changes in shorelinedynamics. In the case of gravel barriers, Carteret al. (1987) identified a number of idealizedmodels of planform response mediated largelyby sediment supply and antecedent topography.The lack of equivalent studies in sandy environ-ments may reflect the more complex dynamicsof those systems. Similarly, the direct relation-ship tentatively identified by Orford et al. (1995)between gravel barrier retreat and sea-level rise is probably masked in sand systems by theseaward return mechanisms and the range ofprocesses operating in the more dynamicallyresponsive sand systems.

Relative sea-level is falling on some temperatecoasts as a result of isostatic uplift, particularlyin paraglacial areas, and also as a result of tecton-ism. Coastal sedimentary responses to fallingsea-level are less well studied than rising sea-level. Typical landforms of regressive coasts arebeach-ridge plains. These lines of shore-parallelbeach-sand ridges are stranded as sea-level falls,isolating earlier formed ridges from the sandsource. Several such plains have been attributedto falling relative sea-level (Dominguez et al.1987; Orford et al. 2003). A series of progradingbeach ridges in north-east Ireland were depositedduring a fall in sea-level between 6000 and 2000yr BP (Orford et al. 2003). Firth et al. (1995)described a series of beaches and spits depositedin an embayment in eastern Scotland during fall-ing sea-level as shelf sand was carried onshore

ESC08 9/7/06 3:20 PM Page 299

300 ANDREW COOPER

by wave action. Falling sea-level enabled wavesto reach sand located progressively further off-shore, and increased the accommodation spacefor sediment accumulation. Littoral deposits,stranded as sea-level falls, may also be subjectedto reworking by wind and blown into dunefields.Many of the extensive coastal sand dunes of thenorth-west Irish coast were emplaced during afall in sea-level between 5000 and 3000 yr BP(Carter & Wilson 1993; Orford et al. 2003).

The understanding of coastline sedimentarybehaviour is of major societal concern becauseof the growing extent and diversity of humaninfluence on the coast. Global climate changeand associated storm patterns, sea-level changesand ecological changes will have an impact onfuture coastal morphology. At present, under-standing of the relationships between forcingfactors and coastline response is, at best, limited.This stems in part from the diverse range of con-trolling factors and in part from heterogeneityin the material factors of the coastline. ManyHolocene coastal deposits have been identifiedthat reflect a rapidly changing environmentalcontext (sea-level and climate change). Thesesequences hold much potential for understand-

ing the likely nature of future coastline changes(e.g. what conditions, of those that determinethe different modes of shoreline response to sea-level change discussed above, are likely to occur?).Problems in their interpretation reflect the par-tial or non-preservation of the full depositionalsequence.

A widespread alternative approach in contem-porary coastal management practice is the useof applied numerical models of shoreline beha-viour. These seek to simulate particular dynamicconditions and to predict future shoreline beha-viour. The appeal of such models lies in theirnormally deterministic output that yields valuesthat can be used in economic planning. The com-plexities of shoreline behaviour outlined here,however, illustrate that such models do not yieldaccurate answers because of their inability totake account of all the important factors.

Planning for sea-level rise frequently involvesthe application of the Bruun concept, which,although it has never been shown to work, hasseen application in at least 26 countries since1995 (Pilkey & Cooper 2004). The applicationof simple models such as this one demonstratesthe need for further investigation of coastline

(a) Offshore dispersal

(b) Vertical accretion

(c) Dune transgression

Deposition SL3-SL1

SL3SL2

Recession SL3-SL1

SL1

Recession SL4-SL3Recession SL3-SL2Recession SL2-SL1

SL3

SL2

SL1

SL1

SL1

Deposition SL1-SL4

SL2

Recession SL2-SL1

Dune erosionErosion

Shelf deposition

Active regressiveforedune growth

TransgressiveDunefield

Fig. 8.19 Conceptual models of coastal duneresponse to rising sea-level. In the erosionalresponse (a) sediment is dispersed offshore. With adequate vegetation, sediment may betransported onshore and trapped in verticallyaccreting foredunes that accrete as seaward dunesare eroded (b). If vegetation cannot withstandsedimentation rates, a transgressive dunefield may form (c). (After Carter 1991.)

ESC08 9/7/06 3:20 PM Page 300


response to sea-level rise in different environ-mental settings. The identification of the role ofsea-level rise is of course difficult to segregatefrom all of the other forcing factors that drivecoastline change at the historical scale.

A key future issue will be to change society’sexpectation of a precise prediction and to accepta qualitative estimate of future shoreline beha-viour based on all available sedimentologicalinformation. Interpretations of future shorelinebehaviour are best derived from all the evid-ence available. This may include stratigraphicalsequences, historical records of morphologicalchange and of driving forces, observations of con-temporary processes and qualitative sedimentarymodels that identify sediment sources, sinks andshoreline behaviour. The ‘Futurecoast’ projectin England and Wales (Cooper & Jay 2002),which adopts such an approach to future plan-ning of the shoreline, is a significant advance onearlier approaches based largely on applicationof numerical models.

Future climate change has the potential to alterspatial and temporal patterns of coastal sedi-mentology. The evidence contained in coastalbeach-ridge sequences outlined above shows theimportance of variations in climatic conditionsat decadal time-scales. Such studies hold muchpotential for understanding the potential role of future climate change, for example increasedperiods of storminess. Again, qualitative estim-ates of future coastal behaviour are likely to bethe best that can be achieved.

Finally, the greatest future threat to temperatecoastlines is the impact of a growing humanpopulation. The potential to have an impact onthe coast in many different ways has been out-lined above. The extent to which these impactsare understood and mitigated will have probablythe most important effect on future coastlines.The human response lies in the realm of politicsand economics as much as science, however,better understanding of coastal sedimentary be-haviour can inform the decision-making process.

ESC08 9/7/06 3:20 PM Page 301

INTRODUCTION

Tropical and subtropical coastlines are charac-terized by a wide array of sedimentary envir-onments including beaches, dunes, deltas andestuaries, and which are broadly comparable totheir temperate counterparts (see Chapters 7 &8). These lower latitude coastlines are also, how-ever, characterized by two unique sedimentaryenvironments, coral reefs and mangroves, andthese, along with their associated sediment sub-strates, are the focus of this chapter. Althoughthe respective environments are characterized byvery different sedimentary processes, sedimentproduction and accumulation in both environ-ments are strongly influenced by biologicalactivity. In coral reefs, the reef structure itself isa product of coral growth, and a high percent-age of the sediment substrate typically derivesfrom the breakdown of calcareous skeletal organ-isms. In mangroves, the trees that colonize theseintertidal settings not only contribute abundantorganic material to the substrate, but also aidsediment trapping and stabilization. As withmost other coastal environments these systems aresubject to a high degree of physical reworking.Such natural sediment and shoreline dynamics,combined with, in places, extensive urbaniza-tion of the tropical coastal fringe, creates aunique set of management challenges. In manyreef and mangrove settings, this is exacerbatedby resource extraction and exploitation. Froman environmental perspective, anthropogenic-related damage to the biological components of both environments has potential knock-on

9Tropical coastal environments:

coral reefs and mangroves

Chris Perry

effects for the physical (including sediment) com-ponents. In mangroves, this relates to sedimentcontamination and the destabilization of sedi-ment substrates following mangrove mortality,and in coral reefs to reduced rates, or modifiedpatterns, of carbonate sediment and frameworkproduction. This chapter examines the sourcesand mechanisms of reef and mangrove sedimentaccumulation, the response of these sediment-ary systems to both natural and anthropogenicchange, issues of shoreline management andremediation, and concludes with a review of thelikely response of coral reefs and mangroves tofuture climatic and environmental change.

9.1 NATURE AND DISTRIBUTION OF CORAL REEF AND

MANGROVE SEDIMENTARY ENVIRONMENTS

Tropical and subtropical regions are characterizedby two distinct sedimentary environments: coralreefs and mangroves. Both are spatially signi-ficant and differ from many other sedimentaryenvironments because of the close links betweenthe ecological aspects of the environments andtheir sedimentology. Tropical coral reefs are, for example, characterized both by the coralsthemselves and by a diverse associated calcare-ous fauna (e.g. calcareous algae, molluscs andforaminifera), which, in combination, contributeto the reef structure and the associated sedi-mentary facies. Similarly, mangrove ecosystemsare defined by a characteristic range of floraland faunal associations, which inhabit inter-tidal sediments and form the basic biological

ESC09 9/7/06 3:21 PM Page 302

TROPICAL COASTAL ENVIRONMENTS: CORAL REEFS AND MANGROVES 303

structure of the mangrove. These mangrove treesand shrubs, in turn, promote sediment trappingand may contribute abundant organic material tothe mangrove substrate. A fundamental featureof both environments is their diversity in rela-tion to geomorphological settings, their spatialand, in the case of reefs, bathymetric extent, andthe sources and rates of sediment productionand accumulation. Change within these systemscan result from both natural and anthropogenicdisturbance. Such changes are capable of alter-ing either the environmental parameters or thesubstrates on which the biological componentsdepend and, as a result, of modifying sedimenta-tion rates.

9.1.1 Distribution and occurrence of coral reefs

Coral reefs are typically associated with shallow,warm, clear tropical and subtropical marine settings (Fig. 9.1). They have been variouslydescribed on the basis of their ecology and geo-logy (see Riegl & Pillar 2000), and are commonly

defined as biologically influenced, wave-resistantbuild-ups of coral framework and carbonatesediment within tropical and subtropical set-tings, and which influence sediment depositionin adjacent areas (Longman 1981; Rosen 1990).Coral communities and coral reefs, however,occur in a far wider range of settings, and form a broader range of structures, than can be con-strained by such traditional definitions. Theyoccur, for example, in a range of more marginalsettings (sensu Perry & Larcombe 2003), whichinclude higher latitude (subtropical to warmtemperate) environments, along with areas influ-enced by high turbidity conditions and cool-waterupwelling (see section 9.1.3). Although theseenvironments may be characterized by reducedrates of in situ calcium carbonate framework andsediment production, and may ultimately producevery different sedimentary deposits, they maystill have high coral cover and thus representimportant sites of coral-related carbonate pro-duction. The coral-related sedimentary environ-ment may thus better be defined as comprising

0 2000 4000

km

60oN

40oN

20oN

0o

20oS

40oS

60oS

60oN

40oN

20oN

0o

20oS

40oS

60oS

60oE 120oE 180o 120oW 60oW 0o

a

b

cd

e

g

h

i

j

kl

m

n

o

p

Mean winter 15°C sst isotherm

f

Mangrove coastlines

Coral reef areas

Mean 20°C winter sst isotherm

12

3

1 - Niger2 - Orinoco3 - Amazon

Cool currents

Warm currents

Fig. 9.1 Global distribution of coral reefs and mangroves compared with mean 15°C and mean 20°C winter sea-surface temperatureisotherms. Latitudinal limits on the distribution of mangroves: a, St Louis, Senegal; b, Lobito, Angola; c, St George, Bermuda; d, StAugustine, Florida; e, Chandeleur Islands, Louisiana; f, Rio Soto La Marina, Mexico; g, Aranangua River, Brazil; h, Puerto de Lobos,Mexico; i, Piura River, Peru; j, Kiire, Kyushu, Japan; k, Raglan Harbour, New Zealand; l, Corner Inlet, Victoria; m, Leschenault Inlet,Australia; n, Qatif, Trucial coast; o, Wadi-Kid, Sinai; p, Kei River, South Africa. (Adapted from Woodroffe & Grindrod 1991; Hubbard 1997.)

ESC09 9/7/06 3:21 PM Page 303

304 CHRIS PERRY

all areas of active coral growth (regardless ofwave resistance or framework-building potential)along with the associated carbonate (or mixedcarbonate:clastic) sedimentary environment (e.g.lagoons, seagrass beds) (Fig. 9.2).

The primary constructional components of areef are the hermatypic corals. As a group theyare defined by their symbiotic relationship withphotosynthetic zooxanthellae algae. These algaeprovide the corals with additional photosyn-thetically derived energy, enabling them to thrivewithin the typically low nutrient waters of thetropics. At a global scale, tropical reef develop-ment can be broadly delineated by the mean 20°Csea-surface temperature isotherm (Fig. 9.1). Thiscorresponds to latitudes between about 28°N and28°S, where reefs currently occupy an estimatedarea of 255,000 km2 (Spalding & Grenfell 1997).Within this latitudinal range, coral growth (andthe potential for reef development) is highlyvariable, and is influenced by a range of factors.These include seawater temperature, aragonite

saturation state, salinity, light and nutrient levels(Table 9.1; Fig. 9.3a). Individual coral speciescan function across wide temperature ranges, butat both higher and lower temperature extremes(Table 9.1) the symbiotic coral–algal relationshipbreaks down and corals respond by shedding theirphotosynthetic algae, with consequent impactsfor coral growth and calcification (Glynn 1996).Temperature also exerts a fundamental controlon calcification because seawater temperature ispositively correlated with aragonite saturationstate (Buddemeier 1997). This influences calciumcarbonate production rates, which decrease withlatitude as sea-surface temperatures decrease(Fig. 9.3b & c).

Corals also survive across a range of salin-ity levels (Table 9.1), but marked reductions inreef-building potential occur in areas subject toeither high fluvial discharge (Fig. 9.1) or intenseevaporation. Given the dependence of hermatypiccorals on photosynthetically derived energy, andin particular the link between photosynthesis and

Fig. 9.2 Atoll reef showing extensive sheets of carbonate sediment (white areas) accumulated on the leeward side of the reef crest(marked by the area of breaking waves), Courtown Cay Atoll, Caribbean. Field of view is approximately 1 km.

ESC09 9/7/06 3:21 PM Page 304

En

cru

stin

g

Rig

id b

ran

che

d

Lim

it of

activ

e co

ral

grow

th

Pho

ticzo

ne

1% s

urfa

ce

light

illu

min

atio

n

Nut

rient

s an

d su

spen

ded

sedi

men

ts

Tem

pera

ture

18 -

25–

29 -

36o C

Sal

inity

22 -

25–

35 -

40

Bac

k-re

ef/la

goon

Ree

f fro

nt

For

e-re

efR

eef

flat

Ree

fcr

est

Coo

l (nu

trie

nt-r

ich)

up

wel

ling

Mou

nd/m

assi

ve

Bra

nche

d Pla

ty

Mou

nd/b

ranc

hed

Sed

imen

t-do

min

ated

sub

stra

teF

ram

ewor

k-do

min

ated

sub

stra

te

Wav

e ba

se

Phy

sica

l di

stur

banc

e

Hig

h-la

titud

e co

ral

reef

s w

ith r

estr

icte

d fr

amew

ork

deve

lopm

ent

2.5

2.0

1.5

1.0

0.5

0.0

CaCO3 accumulation rate (m kyr−1)

Latit

ude

0o 10

o 30

o 4

0o 50

o 60

o

Cor

al-a

lgal

ree

fs.

Ooi

d, c

oral

and

calc

areo

us g

reen

alga

l dom

inat

ed

sedi

men

ts

Cor

allin

e al

gae,

mol

lusc

s,

fora

min

ifera

(co

ral a

nd

calc

areo

us g

reen

alg

ae

pres

ent,

but l

ess

impo

rtan

t) Cor

allin

e al

gae,

m

ollu

sc a

nd

bryo

zoan

do

min

ated

se

dim

ents

Ara

gonite C

alc

ite

4.5

4.0

3.5

3.0

Aragonitesaturation factor

30

25

20

15

10 5

Sea-surfacetemperature (oC)

Latit

ude

0o 10

o 2

0°

3

0o 4

0o 50

o 60

o

(a)

(b)

(c)

Fig.

9.3

(a) E

nviro

nmen

tal c

ontro

ls o

n th

e de

velo

pmen

t of c

oral

reef

com

mun

ities

. (A

dapt

ed fr

om Ja

mes

& B

ourq

ue 1

994.

) (b)

Lat

itudi

nal c

hang

es in

CaC

O3

accu

mul

atio

n ra

tes

and

(c) s

ea-s

urfa

ce te

mpe

ratu

re a

nd a

rago

nite

satu

ratio

n st

ates

. (A

dapt

ed fr

om B

udde

mei

er 1

997;

sedi

men

t typ

es a

fter L

ees 1

975.

)

ESC09 9/7/06 3:21 PM Page 305

306 CHRIS PERRY

coral calcification, light is a key control on coralgrowth. Light decreases with depth, so that ratesof coral growth and calcification also decrease(Huston 1985). The lower limit of hermatypiccoral growth is defined as the base of the photiczone (where surface light levels are reduced to1%; Fig. 9.3a). In clear-water settings this can beas deep as around 90 m (Table 9.1), but occursat much shallower depths in turbid environmentsdue to reduced light penetration. Finally, elevatednutrient levels (nitrate levels > 2.0 μmol L−1;phosphate levels > 0.20 μmol L−1) may result in reduced rates of coral growth (Tomascik &Sander 1985).

9.1.2 Distribution and occurrence of mangroves

The term mangrove is variously defined and has been used to refer either to the constituentplants of these tropical intertidal forests or toboth the community and its associated envir-onment (Tomlinson 1986). The main definingcharacteristics of mangroves are that they com-prise communities of salt tolerant tropical/sub-tropical trees and shrubs, and as such representtropical equivalents of temperate salt-marshcommunities (Woodroffe 1983). In this chapter,mangroves are discussed not only in the con-text of the constituent trees and shrubs, but alsothe sediment substrates on which the mangrovesdevelop and the creek networks that dissect

them (Fig. 9.4). Mangrove ecosystems extendalong some 60–75% of tropical and subtropicalcoastlines (MacGill 1958) and recent mappingestimates suggest a global coverage of around190,000 km2 (Spalding et al. 1997).

Opinions vary about the environmental factorslimiting mangrove development, with both meanminimum air and sea temperatures having beencited as controls (Woodroffe & Grindrod 1991).Some authors also cite the occurrence of extremeseasonal cold (frost) events (Plaziat 1995). Atthe global scale, mangrove distributions exhibita reasonably close correlation with the meanwinter 15°C sea-surface isotherm (Woodroffe &Grindrod 1991), which equates to a latitudinalrange between about 30°N and 30°S (Fig. 9.1).Actual distributional patterns, however, are vari-able and reflect local environmental (particularlycool seasonal temperature) constraints, such that(as with coral reefs) their distribution is morerestricted on the western coasts of Africa andAmerica (Plaziat 1995; Fig. 9.1).

Such climatic and/or oceanographic constraintsare exacerbated by physical constraints, such aslocal geomorphology, tidal range, seasonal hydro-logy and substrate availability for colonization(Fig. 9.5). These factors influence mangrovedevelopment both at local and regional scales.At the local scale marked physical–chemicalgradients across shorelines are determined by therelative importance of tidal against freshwater

Table 9.1 Marine environmental parameters that influence the distribution of hermatypic corals and of tropical coral reef development.‘Optimal’ values for coral growth are shown, along with recorded upper and lower environmental limits. Figures in parentheses are fornon-reef building coral communities. (Data from Kleypas et al. 1999.)

Environmental parameter ‘Optimal’ levels Environmental limits

Lower Upper

Temperature (°C)* 21.0–29.5 16.0 (13.9) 34.4 (32.1)Salinity (PSU)† 34.3–35.3 23.3 (20.7) 41.8 (No data)Nitrate (μmol L−1)‡ < 2.0 0.00 3.34 (up to 5.61)Phosphate (μmol L−1)‡ < 0.2 0.00 0.40 (up to 0.54)Aragonite saturation state (Ω-arag)§ c. 3.83 3.28 (3.06) No dataDepth of light penetration (m) c. 50 < 10 c. 90

*Weekly data.†Monthly average data.‡Overall averages (1900–1999).§Overall averages (1972–1978).

ESC09 9/7/06 3:21 PM Page 306


Fig. 9.4 Fringing mangrove developed along the shoreline of Inhaca Island, Mozambique. Note the well-developed creek systems thatdissect the mangrove. Field of view approximately 0.5 km. (Photograph courtesy of Simon Beavington-Penney.)

HWS

HWN

Avicennia Rhizophora Bruguiera Ceriops Xylocarpus

Tidalrange

Substrate composition, e.g. ratio of sand/mud/clay; carbonate vs clastic content

Fluvial and/orgroundwater

influence

Seawardzone

Mesozone Landwardzone

Terrestrial zone

Mean air temperature>10–19o C (or absenceof frosts)

Sedimentation rateMean sea surface temperature >15o C

Rainfall/aridity/seasonality

Shoreline geomorphology

Salinity gradient

Fig. 9.5 Environmental controls on the development and zonation of mangroves: HWS, high water, spring tides; HWN, high water,neap tides. Temperature ranges are based on data in Woodroffe & Grindrod (1991).

ESC09 9/7/06 3:21 PM Page 307

0 10 20 30 40 50 60 70

Depth (m)

(a) W

arm

, cle

ar-w

ater

set

tin

g

Thi

ck (

5–10

m)

cora

ldo

min

ated

fram

ewor

k

Ree

f in

itiat

ion

0 5 10 15Depth (m)

(b) T

urb

id s

etti

ng

Non

-fra

mew

ork

cora

l co

mm

uniti

es

0 10 20 30 40

Depth (m)

(d)

Up

wel

ling

-in

flu

ence

d s

etti

ng

0 5 10 15

Depth (m)

(c)

Hig

h la

titu

de

sett

ing

Det

rital

cor

al r

ubbl

e fra

mew

ork

set w

ithin

mix

ed c

arbo

nate

:cla

stic

sed

imen

ts

Cor

al g

row

th s

patia

lly

and

bath

ymet

rical

ly r

estr

icte

d

e.g.

Inha

ca Is

land

, Moz

ambi

que

(Per

ry 2

003)

Insh

ore

cent

ral G

reat

Bar

rier

Ree

f (S

mith

ers

& L

arco

mbe

200

3)

Sof

t cor

al-d

omin

ated

co

mm

unity

Non

-fra

mew

ork

cora

l com

mun

ity

e.g.

Nat

al c

oast

, S. A

fric

a (R

iegl

et a

l. 19

95)

e.g.

Gul

f of A

den,

Yem

en (

Ben

zoni

et a

l. 20

03)

Hig

h or

low

rel

ief

cora

l car

pets

Und

erly

ing

bedr

ock/

sedi

men

t sub

stra

tes

Cor

al fr

amew

ork

Cor

al c

olon

ies/

carp

et c

omm

uniti

es

e.g.

nor

th J

amai

ca (

Lidd

ell &

Ohl

hors

t 198

8)

Hig

h co

ral c

over

ext

endi

ng

to d

epth

s of

60–

70 m

Sea

sona

l co

ol

upw

ellin

g

Sed

imen

tre

susp

ensi

on/

turb

idity

Per

iodi

c st

orm

dist

urba

nce

Fig.

9.6

Sim

plifi

ed re

pres

enta

tions

of c

oral

fram

ewor

k de

velo

pmen

t in

a ra

nge

of e

nviro

nmen

tally

lim

ited

setti

ngs.

(Ada

pted

from

Per

ry &

Lar

com

be 2

003.

)

ESC09 9/7/06 3:21 PM Page 308


influence, as well as by the seasonality of rainfall(Fig. 9.5). Substrate salinity levels are particularlyimportant and mangrove colonization is enhancedwhere freshwater inputs, from either river orgroundwater sources, dilute salinity levels. At theregional scale, different climatic regimes result invariable fluvial runoff and substrate availability(Tomlinson 1986). Such variable influences resultin complex and diverse sediment transport andaccumulation processes in different settings, andthe development of a diverse range of mangrovesedimentary environments.

9.1.3 Reef types and geomorphology

Reef structures are commonly described in termsof their overall geomorphology and proximity toadjacent landmasses. James & Macintyre (1985)delineate five basic reef types: (i) fringing reefs,(ii) bank-barrier reefs, (iii) barrier reefs, (iv) atollsand (v) patch reefs. The extent of reef develop-ment at individual sites is, however, highly vari-able and reflects local environmental parameters(see section 9.1.1). In warm, clear water and low nutrient environments, reefs can form bothspatially and bathymetrically extensive struc-tures. This state is exemplified by the fringingreefs of north Jamaica, where coral growth andframework development has occurred to depthsof 80+ m (Liddell & Ohlhorst 1988; Fig. 9.6a)and where framework structures some 5–10+ mthick have developed during Holocene times(Land 1974). In contrast, restricted coral com-munities, with limited framework development,occur in turbid, nearshore environments. AroundInhaca Island, southern Mozambique, coralcommunities are restricted by low light levels todepths of < 6 m (Perry 2003) and frameworkdevelopment is replaced by unconsolidated coralrubble, set within a carbonate:clastic sedimentmatrix (Fig. 9.6b).

Marked variations in calcium carbonateaccumulation rates (and reef development) alsooccur as mean sea-surface temperatures andaragonite saturation state decrease (Fig. 9.3b;Buddemeier 1997). As a result, reef frameworkdevelopment typically decreases towards higherlatitude areas as environmental conditions

become progressively more marginal for coralsurvival, and carbonate accumulation ratesinsufficient for framework construction. Thisstate is illustrated by the reefs along the Natalcoast of South Africa (Riegl et al. 1995), wherecoral communities colonize subtidal bedrock,but framework accumulation does not occur(Fig. 9.6c). Similar reductions in coral growthpotential can occur in lower latitude settings (a ‘pseudo-high-latitude effect’; Sheppard &Salm 1988) where seasonal upwelling brings cool,nutrient-laden waters to the surface, for examplethe Gulf of Aden (Fig. 9.6d; Glynn 1993). Atthese sites, framework development can be bothspatially and bathymetrically restricted. Suchexamples illustrate the diverse range of environ-mental settings in which coral reefs occur, andwhich in turn influence patterns of reef sedimentaccumulation.

9.1.4 Mangrove types and geomorphology

Although mangroves are typified as formingextensive swamps associated with shorelines andestuaries that are accumulating sediment, theyactually occupy a diverse range of coastal, off-shore (island) and fluvially influenced settings(Thom 1982; Woodroffe 1992). These include:1 alluvial plains – areas characterized by highfluvial sediment accumulation, throughput ordischarge (Fig. 9.7a);2 tidal plains – areas of high tidal range charac-terized by strong bi-directional flow patterns(Fig. 9.7b);3 wave-protected coastlines – mangroves developalong the landward sides of barrier islands andbeach ridges (Fig. 9.7c), and the shorelines ofprotected lagoons (Fig. 9.7d);4 coastal embayments and drowned valleys(Fig. 9.7e);5 carbonate-dominated coastal environments –these include subtidal carbonate mudbanks, andsubstrates associated with intertidal reef flats(Fig. 9.7f).

Although mangroves thus occur in a range of geomorphological settings, they also occuracross a range of climatic settings. These spanarid, through subtropical to tropical shorelines.

ESC09 9/7/06 3:21 PM Page 309

(a)

Del

taic

set

tin

ge.

g. F

ly R

iver

Del

ta

(Wol

ansk

i et a

l. 19

98)

a

a1

Aba

ndon

edde

ltaic

pla

in Act

ive

delta

icpl

ain

aa1

Man

grov

e

Nea

rsho

re/m

angr

ove

mud

s

Flu

vial

san

ds

Flo

odpl

ain

mud

Coa

stal

san

d

MH

W

(b) T

idal

set

tin

ge.

g. N

orm

anby

Riv

er, A

ustr

alia

(Bry

ce e

t al.

1998

)

Est

uarin

epl

ain

Upl

and

Riv

erLe

vee

b1

b

b1b

Allu

vial

pla

in

MH

W

(c) W

ave-

pro

tect

ed s

etti

ng

e.g.

Can

anei

a la

goon

, Bra

zil

(Ant

hony

& B

livi 1

999)

Lago

on

Bar

rier

Tid

al

delta

c c1

cc1

MH

W

Bar

rier

d1d

Lago

on

Aba

ndon

eddi

strib

utar

ies

Bar

rier

beac

h

(d)

Riv

er/w

ave

sett

ing

e.g.

Grij

alva

Del

ta, M

exic

o(T

hom

196

7)

dd1 M

HW

Lago

onLe

vee

e1

e

Allu

vial

flood

pla

in

(e)

Dro

wn

ed b

edro

ck v

alle

ye.

g. D

arw

in H

arbo

ur, A

ustr

alia

(Woo

drof

fe e

t al.

1988

)

MH

We1

e

Ree

f cre

stSan

dca

y

Sto

rmrid

ge

f

f1

(f)

Car

bo

nat

e se

ttin

ge.

g. L

ow Is

les,

Aus

tral

ia(S

todd

art 1

980)

Hol

ocen

e re

ef

HW

M

f1f

Ree

f cre

st

Fig.

9.7

Man

grov

e de

posi

tiona

l set

tings

: HW

M, h

igh-

wat

er m

ark;

MH

W, m

ean

high

wat

er. (

Ada

pted

from

Woo

drof

fe 1

992.

)

ESC09 9/7/06 3:21 PM Page 310


These may be characterized by different man-grove species depending upon species toleranceto environmental stress. Avicennia, for example,are most resistant to low and high temperatures,and to high soil salinities, whereas Rhizophoraare resistant to low temperatures (Plaziat 1995).Along arid shorelines, the major environmentalgradients relate to salinity, which increaseslandward due to high evaporation rates. In these environments mangroves thus often form narrow fringes of dwarf Avicennia (< 1 m high).In contrast, dense, sprawling forests develop onlarge areas of deltaic mud and organic-rich sub-strates along subequatorial shorelines (Smith1992). In these cases, the dominant Rhizophoraand Avicennia plants may reach heights of 40 m(Plaziat 1995).

9.2 SEDIMENT SOURCES AND SEDIMENT ACCUMULATION

PROCESSES

9.2.1 Sources and characteristics of coral reefsediments

Sediments that accumulate on and around coralreefs derive from a range of sources. Theseinclude:1 skeletal sediments – the calcareous remains of reef framework-building and reef-associatedorganisms;2 non-skeletal sediments – grains produced byphysico-chemical induced carbonate precipita-tion;3 allochthonous sediments – grains derivedfrom terrestrial sources and which may be eithernatural or anthropogenic in origin.The relative abundance of these sediment contributors varies both within and betweenenvironments.

9.2.1.1 Skeletal sediments

Typically the most abundant constituent of reef-related sediments are the skeletal remnantsof calcareous reef organisms. These can be sub-divided into sediments produced by the break-down of carbonate framework contributors,

and those derived from reef-associated benthicorganisms and calcareous algae. Coral reefframework is composed of two main construc-tional components, the skeletons of hermatypiccorals (primary framebuilders) and a diversearray of associated calcareous encrusting faunas(secondary framebuilders). Calcareous encrusters(crustose coralline algae, encrusting forms ofbryozoans and foraminifera, and serpulids) pro-duce multiple crusts on the dead surfaces of coralskeletons (Martindale 1992) and help bind andstabilize the reef framework (Rasser & Riegl2002). Breakdown of these primary and secondaryframework contributors (and thus framework-related sediment production) is facilitated byphysical and biological activity.

Physical (storm) disturbance results in frag-mentation and transport of coral framework,and generation of coral rubble (Fig. 9.8), whichin turn can be degraded by physical reworking toproduce fine coral sand/silt (Fig. 9.9). However,the release of framework carbonate into sedi-ment results primarily from bioerosion (a termused to describe biological substrate erosion;Neumann 1966). Bioerosion is facilitated by a wide range of reef-associated faunas, includ-ing fish and echinoids, and endolithic forms ofsponges, bivalves and worms (Fig. 9.8; Hutchings1986). Framework degradation and sedimentproduction by fish and echinoid species results asa by-product of the search for food. Parrotfishand surgeonfish, for example, have heavily cal-cified mouthparts and bite off chunks of coralsubstrate, which is excreted as fine sand (Fig. 9.9;Gygi 1975). Similarly, echinoids such as Diademasp. have heavily calcified feeding apparatusenabling them to remove coral skeleton duringfeeding (Fig. 9.8). As a by-product, they pro-duce abundant carbonate-rich faecal pellets(Fig. 9.9; Scoffin et al. 1977).

Significant degradation of framework alsoresults from the activities of endolithic boringorganisms. These include specific groups ofsponges, bivalves and worms (Bromley 1978;Perry 1998a). These organisms, which use eitherphysical and/or chemical processes to excavatetunnels/chambers within dead coral skeleton,produce boreholes > 1 mm in diameter and are

ESC09 9/7/06 3:21 PM Page 311

Car

bona

te n

eedl

esre

leas

ed d

urin

g br

eakd

own

of c

alca

reou

s gr

een

alga

e(P

enic

illus

, Udo

tea)

Bre

akdo

wn

ofH

alim

eda

Inpu

ts o

f cal

care

ous

epib

iont

s (c

oral

line

alga

e,

fora

min

ifera

, ser

pulid

s)

Infa

unal

/epi

faun

al

skel

etal

inpu

ts (

mol

lusc

s,ec

hnio

ids)

Infa

un

al/e

pif

aun

alse

dim

ent

con

trib

uto

rs

Bre

akd

ow

n o

f co

ral

fram

ewo

rk a

nd

co

ral

rub

ble

Bre

akdo

wn

of

calc

areo

us

gree

n al

gae

(Hal

imed

a)In

tern

al b

ioer

osio

n (s

pong

es, b

ival

ves,

w

orm

s)

Fis

h gr

azin

g re

leas

es fi

neca

rbon

ate

sedi

men

t

Rel

ease

of f

ragm

ents

of

cal

care

ous

encr

uste

rs(c

oral

line

alga

e,

bryo

zoan

s, fo

ram

inife

ra)

Win

now

ing

of

bouy

ant g

rain

s

Ons

hore

/offs

hore

tran

spor

t of

sedi

men

t dur

ing

stor

ms

Lago

on s

edim

ents

mov

ed

seaw

ard

durin

g st

orm

s

Alg

al m

ats

stab

ilize

su

bstr

ates

Sed

imen

t set

tling

and

trap

ping

with

in s

eagr

ass

beds

Dee

p se

dim

ent

biot

urba

tion

by s

hrim

ps

Sed

imen

t rew

orki

ng b

yep

ifaun

al d

etrit

ivor

es(H

olot

huria

ns)

Ter

rigen

ous

sedi

men

t an

d an

thro

poge

nic

part

icul

ate

inpu

ts

Ech

inoi

d gr

azin

g pr

oduc

es

faec

al c

arbo

nate

pel

lets

Ool

itic

sand

form

atio

n

Long

shor

e se

dim

ent

inpu

ts/o

utpu

ts

Tra

nspo

rt/r

ewor

king

pat

hway

s

Pro

duct

ion

path

way

s

Sed

imen

t sco

urin

g du

ring

stor

ms

Fig.

9.8

Sche

mat

ic d

iagr

am il

lust

ratin

g th

e m

ain

sedi

men

t sou

rces

and

tran

spor

t pat

hway

s with

in c

oral

-reef

-rela

ted

envi

ronm

ents

.

ESC09 9/7/06 3:21 PM Page 312


termed macroborers. In addition, a range ofmicroborers, including forms of cyanobacteria,chlorophytes and fungi, facilitate substrate de-gradation on a microscopic scale (Golubic et al.1975; Perry 1998b). Although all boring activ-ity results in framework degradation, not allresults in sediment production, because varyingproportions of the excavated material will bedirectly dissolved. The most significant sedimentproducers are the boring sponges. These spongesexcavate small fragments of coral skeleton, whichare subsequently expelled (Fig. 9.8), resulting in the production of abundant fine-grained sand(Fig. 9.9; Fütterer 1974). Once released into thesediment carbonate grains are subject to fur-ther physical, chemical and biological alteration(Perry 2000).

Coral framework is not, however, the onlysource of reef sediment. Skeletal carbonate sedi-ment is also derived from infaunal and epifaunalorganisms as well as a wide array of calcifying

algae. Common infaunal and epifaunal sedimentcontributors include bivalves and gastropods,foraminifera and echinoids (Fig. 9.8; Swinchatt1965). Skeletal sediments are also contributedby calcareous epiphytic organisms (includingcrustose coralline algae, foraminifera, serpulidsand bryozoans) that encrust seagrass blades andthe stems of calcifying algae (such as Udoteaand Penicillus; Fig. 9.8; Land 1970; Nelson &Ginsburg 1986). In addition, carbonate sedimentis produced during breakdown of carbonate-secreting algal species. Some, such as Penicillusand Udotea, release abundant carbonate needlesinto the sediment (Neumann & Land 1975),whereas others, such as Halimeda, produce heav-ily calcified segments, which often subsequentlydisintegrate into carbonate needles (Fig. 9.9).Many calcifying algae have high turnover rates,producing several standing crops per year, andare thus significant components of reef sedi-ment budgets (Neumann & Land 1975). Minor

Cliona spongeboring

Parrotfish

Diademaurchin

Halimeda

Coral

Mechanical breakdown

Biological breakdown

Segments(4–8 mm)

Broken segments(1 mm)

Dust (1 µm)

Joints (64 mm)Grit (250 µm)

(< 63 µm)

Coral chips

(500–125 µm)

Faecal carbonate

(1 mm–250 µm)

Faecal pellets

NB. Most particles are also subject to further diminution by microendolithic borers

(250–125 µm)

Coralfragments

256 64 16 4 1 250 63 15.6 3.9 0.9mm µm

256 64 16 4 1 250 63 15.6 3.9 0.9mm µm

)

Fig. 9.9 Mechanical and biologicalbreakdown of carbonate sedimentcontributors and resultant sediment sizefractions produced. (Adapted fromScoffin 1987.)

ESC09 9/7/06 3:21 PM Page 313

314 CHRIS PERRY

additional skeletal, but non-carbonate, sedimentsderive from diatoms and sponge spicules (both ofwhich are formed from silica). Morphologicaldescriptions of these varied sediment contributorsare found in Scoffin (1987).

Production rates by different contributinggroups and processes vary markedly between reefenvironments, thus dictating spatial variationsin rates of sediment production and accumula-tion. At Kailua Bay (Oahu, Hawaiian Islands),both total carbonate sediment production ratesand the production rates of specific sedimentcontributors vary between environments. Near-shore hardgrounds have the highest sedimentproduction rates (0.62 kg m−2 yr−1) with produc-tion dominated by Halimeda (0.25 kg m−2 yr−1),bioerosion-derived coral (0.13 kg m−2 yr−1) andmolluscs (0.18 kg m−2 yr−1). In contrast, sedi-ment production rates in reef front sites average0.41 kg m−2 yr−1, of which 95% is derived fromcoral bioerosion (Harney & Fletcher 2003).

9.2.1.2 Non-skeletal carbonate sediments

In addition to skeletal sediments, non-skeletalgrains also occur within reef and reef-relatedsediments. These include carbonate muds, peloidsand ooids (Fig. 9.8). Carbonate silts (grain size < 63 μm) often comprise a volumetrically signi-ficant component of lagoon sediments. Much of this material comprises single aragonite crys-tals derived from the breakdown of carbonate-secreting marine algae, although some may alsoresult from direct precipitation from super-saturated sea waters. Peloids are small, roundedor elliptical grains characterized by a micro-crystalline internal structure and have a range of origins (Macintyre 1985). Some represent thecalcified remnants of faecal pellets, others frag-ments of skeletal grains where the internal struc-ture has been modified by microboring. Somepeloids are also believed to result from phases ofchemical precipitation around a central nucleus.Ooids are also small rounded grains (typically < 1 mm diameter), but exhibit multiple concen-tric lamellae that form a coating around a centralnucleus (often a peloid or skeletal fragment) andresult from physico-chemical precipitation of

calcium carbonate under high-energy conditions(Ginsburg 1957).

9.2.1.3 Allochthonous sediments

Although the majority of sediment that accumu-lates on and around coral reefs is carbonate,there are significant external inputs of sedimentin some environments. This is particularly evid-ent in areas where reefs develop close to sourcesof fluvial sediment discharge (Fig. 9.6). Typic-ally the proportion of terrigenous sediment willdecrease with distance from source (e.g. Acker& Stearn 1990), although sediment accumula-tion within individual coastal environmentscommonly reflects not only local carbonate andfluvially derived sediment inputs, but also sedi-ment flux into and out of the environment. A detailed sediment budget of a carbonateembayment at Hanalei Bay, Hawaii, influencedby terrigenous sediment input (Calhoun et al.2002), found that significant proportions of fluvi-ally derived suspended sediments were exportedoffshore, and the bay was also subject to inputs of carbonate sediment derived from adjacentcoastal areas. In this example, the total volumeof Holocene carbonate sediment that has accu-mulated in the bay exceeds estimates of in situHolocene carbonate sedimentation, and the bayis therefore acting as a net sink for sediment produced in adjacent coastal areas. Increasinglyassociated with terrigenous sediment inputs are a range of dissolved and particulate con-taminants linked to anthropogenic (industrialor agricultural) discharges. These include heavymetal and hydrocarbon contaminants and arediscussed in section 9.4.4.

9.2.2 Controls on coral reef sediment transportand accumulation

Although local environmental factors influencethe composition and abundance of individualsediment contributors, the accumulation of carbonate sediment in reef environments is influ-enced by a wide range of physical and biogenicprocesses. These influence sediment transport,reworking, trapping and stabilization.

ESC09 9/7/06 3:21 PM Page 314


9.2.2.1 Reef sediment transport

Sediment transport and deposition are deter-mined by two main factors: (i) shear stress and(ii) settling velocity. The former relates to thevelocities required to move or entrain sedimentparticles of a specific size, the latter to the differ-ence between the gravitational and buoyancyforces acting on the particle. The relationshipbetween threshold and settling velocities is rela-tively well established for quartz grains, which,due to uniform densities, behave in a reasonablypredicable fashion dictated by grain size (seeChapter 1). Grain transport and deposition incarbonate sediments, however, are complicated bydifferences in grain skeletal structure (and hencedensity) and by grain size, shape and texture.Grains with plate-like morphologies will settleat a slower rate than block or rod-shaped grainsand hence such parameters influence grain trans-port and deposition (Kench & McLean 1996).

These hydraulic controls have been well illus-trated in a study of sediment transport in theCocos (Keeling) Islands (Kench 1997). Carbonatesediment assemblages around the atoll can berelated to classes defined by settling velocitiesand which delineate sediment transport path-ways. Essentially two main sediment assemblagesare present: one dominated by reef-derived com-ponents, and a second produced in situ withinthe lagoon. The former occur within what hasbeen described as an ‘active transport zone’ in which reef-derived sediments are selectivelytransported by currents from shallow reef flatareas into the channels entering the centrallagoon. Sediments in these areas are dominatedby faster settling grains (mainly larger fragmentsof coral, coralline algae and the foraminiferaAmphistegina sp.). Slower settling grains (smallercoral grains, Halimeda and the foraminiferaMarginopora sp.) are transported through thereef flat channels and deposited along the lagoonmargins (Kench 1997).

9.2.2.2 Reef sediment trapping and stabilization

Although the physical properties of grains dic-tate sediment entrainment and transport, a range

of biogenic components and physico-chemicalprocesses interact to trap and stabilize reef sedi-ments. Particularly important in this respect are marine seagrasses and green algae. The longblades of seagrasses, such as Thalassia andSyringodium, locally reduce current speeds andpromote sediment settling (Scoffin 1970). In the long-term, such processes can lead to thedevelopment of carbonate mudbanks (Bosenceet al. 1985), although this will depend uponlocal rates of carbonate mud production (Perry& Beavington-Penney 2005). The dense rhizome(root) networks associated with seagrasses alsofacilitate substrate stabilization and binding.Similar binding of sediment occurs around theholdfasts of green algae such as Halimeda,Penicillus and Udotea (Scoffin 1970). Organicbinding of sediment also occurs in areas wherealgal-mat communities develop (most commonlyin low-energy lagoon settings). Associated withthese mats are filamentous algae including thecyanobacteria Lyngbya and Schizothrix, and thechlorophyte Enteromorpha (Scoffin 1970), whichpromote sediment adhesion and trapping, and inturn enhance substrate stabilization. Binding ofsubstrates also occurs in areas subject to physico-chemical and organically induced carbonatecement precipitation (Scoffin 1987).

9.2.2.3 Reef sediment reworking

In addition to sediment stabilization, significantsediment reworking also occurs, much of whichcan be attributed to wave and current action (seesection 9.2.2.1) and to grain diminution (seesection 9.2.1.1). Bioturbation also occurs, how-ever, associated with surface feeders such as holo-thurians and crabs, and subsurface organismssuch as shrimps. Holothurians, for example, areestimated to ingest and excrete up to 250 g ofsediment per day. The excreted sediment not onlyproduces a highly homogenized surficial sedimentlayer, but the ingestion process may also resultin chemical grain dissolution (Hammond 1981).Extensive sediment reworking (with volumes ofsediment turnover in excess of 11 kg m−2 yr−1)may occur associated with infaunal organismssuch as the Callianassa shrimp (Bradshaw 1997).

ESC09 9/7/06 3:21 PM Page 315

316 CHRIS PERRY

Such intense bioturbation results in preferentialsediment sorting and aeration of surface sedi-ments, and creates a highly mobile surface layerpoorly conducive to colonization (Tudhope &Scoffin 1984). Finer sediments expelled duringburrowing are also prone to resuspension andtransport (Roberts et al. 1981).

9.2.3 Spatial variations in sediment accumulation

Spatial variations in the abundance and pro-ductivity of reef sediment contributors, and the processes of sediment transport and reworking,produce distinct grain assemblages in differentreef areas. Across nearshore fringing or bank-barrier reefs, local variations in sediment typesare evident between lagoon, reef crest, and shallow and deep reef front environments. Eachsubenvironment can be delineated on the basisof grain assemblage (i.e. the relative abundanceof skeletal grain types) and texture. These reflectnot only initial grain inputs, but also grainreworking and transport. Across the narrow (c. 1 km wide) fringing reefs of north Jamaica,clear patterns of sediment accumulation can beidentified (Fig. 9.10a), which broadly reflect the abundance of sediment contributing groupson the reef (Boss & Liddell 1987). Lagoon sedi-ments are, for example, characterized by rela-tively high abundances of Halimeda, benthicforaminifera and molluscs, whereas shallowreef-front sediments are dominated by coral,coralline algae and encrusting foraminifera.These patterns are representative of sedimentaccumulation patterns across many narrow shelfreefs. Distinct patterns of carbonate sedimentaccumulation are also evident across larger car-bonate shelf or platform environments. Reefsediment assemblages (comparable to thosedescribed above) occur where reefs are developedalong the seaward margins, but significant car-bonate production can also occur across theinner platform areas where oolitic sand bodies,seagrass beds or green algal meadows develop(Purdy 1963). Such large-scale patterns of car-bonate sediment accumulation are evident acrossthe Great Bahama Bank (Fig. 9.10b), and theland-attached Florida and Great Barrier Reef

shelf settings. Detailed sedimentological descrip-tions of the two former examples are found inTucker & Wright (1990).

Although clear patterns of cross-reef or cross-shelf sediment facies therefore can be identified,marked variations in grain assemblages are alsoevident between reef settings. As outlined in section 9.1.3, the extent of reef developmentvaries markedly between regions and is influencedby spatial and latitudinal changes in marineenvironmental parameters such as tempera-ture, light penetration and aragonite saturationstate. These same limiting factors also influencepatterns and rates of reef sediment productionand are clearly illustrated in relation to shifts inthe types of skeletal and non-skeletal sedimentcontributors. On a global scale these have beenrelated to latitudinal changes in temperatureand salinity (Lees 1975), although related shiftsin aragonite saturation state (Buddemeier 1997;Fig. 9.3) are also important. Lees (1975) iden-tified three distinct carbonate grain assemblagesthat characterize different temperature andsalinity zones:1 a chlorozoan assemblage (characterizing trop-ical waters) dominated by corals and calcareousgreen algae;2 a chloralgal assemblage (characterizing sub-tropical waters) in which corals and calcareousgreen algae become progressively less commonand the dominant grain types are coralline redalgae, molluscs and foraminifera;3 a foramol association (characterizing warmtemperate to cold waters) in which coralline redalgae, molluscs and bryozoans dominate.Transitions between such assemblages there-fore occur as environmental conditions becomeprogressively more marginal for coral survival(Halfar et al. 2000).

9.2.4 Sources and characteristics of mangrove sediments

Mangroves colonize a wide range of coastalenvironments (see section 9.1.4) and the sedi-ments that contribute to mangrove substratesare derived from a range of sources. These canbe classified as either:

ESC09 9/7/06 3:21 PM Page 316

200

m 1

00 m

20 m

40 m

60 m

80 m

Shor

ezo

ne

Lago

on Z

one

Rear

zone

Barre

nzo

neM

ixed

zone

A. c

ervic

orni

s zo

ne

ESCA

RPM

ENT

Pinn

acle

s

Fore

-reef

slope

Deep

fore

-reef

40 20 0

%

Cora

l

Cora

lline

Alg

ae

Halim

eda

Encr

ustin

g fo

ram

inife

ra

Bent

hic

fora

min

ifera

Mol

lusc

Bim

ini

Isla

nds

Joul

ters

Cay

s

ANDROS

ISLAND

Tongue of the Ocean

200

m

200m

0

k

m

40

N

(a)

Cor

al r

eefs

Cor

al a

lgal

Ool

itic

Gra

pest

one

Mud

Pel

let m

ud

(b)

Fig.

9.1

0Sp

atia

l var

iatio

ns in

sedi

men

t con

tribu

tors

and

sedi

men

t fac

ies.

(a) A

cros

s the

nar

row

frin

ging

reef

s of n

orth

Jam

aica

. (A

dapt

ed fr

om B

oss &

Lid

dell

1987

.) (b

) Acr

oss t

heA

ndro

s Pla

tform

, Gre

at B

aham

a Ba

nk. (

Ada

pted

from

Pur

dy 1

963.

)

ESC09 9/7/06 3:21 PM Page 317

Oys

ters

For

amin

ifera

Gas

trop

ods

Biv

alve

sR

apid

dis

solu

tion

ofsk

elet

al m

ater

ial i

n ac

idic

por

e w

ater

s

Infa

un

al/e

pif

aun

alse

dim

ent

con

trib

uto

rs

Org

anic

litte

rpr

oduc

tion

(leav

es, t

wig

s,br

anch

es)

Gro

undw

ater

flow

influ

ence

s so

il ch

emis

try

Dec

ompo

sitio

nof

org

anic

mat

eria

lby

det

ritiv

ores

and

bact

eria

Bio

turb

atio

n by

crab

s, w

orm

s,cr

usta

cean

s

Iron

/car

bona

teno

dule

s fo

rmun

der

oxid

izin

gco

nditi

ons

Sal

twat

er

flocc

ulat

ion

ofcl

ay/s

ilt

part

icle

s

MH

W

MLW

Sec

onda

ry c

ircul

atio

n pr

omot

es fl

ow s

epar

atio

n –

fine

sedi

men

ts d

epos

ited

on u

pper

ban

ks

Ban

ker

osio

nS

heet

er

osio

n

Flo

od-t

ide

tran

spor

t of

susp

ende

d se

dim

ents

Sed

imen

t tra

ppin

g by

mic

robi

al m

ats

on s

oil/r

oots

Pel

letiz

atio

n of

sed

imen

t by

det

ritiv

ores

Set

tling

of f

locc

ulat

edse

dim

ent p

rom

oted

by

vege

tatio

n-in

duce

d fr

ictio

nMec

hani

cal/c

hem

ical

deflo

ccul

atio

n pr

omot

esse

dim

ent e

xpor

tS

uspe

nded

sed

imen

t ad

here

s to

muc

us o

n w

ater

sur

face

-exp

orte

d by

ebb

-tid

es

Flo

od-

and

ebb-

tide,

an

d flu

vial

driv

en

sedi

men

t tra

nspo

rt

Tran

spor

t/rew

orki

ng p

athw

ays

Pro

duct

ion

path

way

s

Incr

ease

d w

ater

flow

thro

ugh

burr

ows

Fig.

9.1

1Sc

hem

atic

dia

gram

illu

stra

ting

the

mai

n se

dim

ent s

ourc

es a

nd tr

ansp

ort p

athw

ays w

ithin

man

grov

e en

viro

nmen

ts. M

HW

, mea

n hi

gh w

ater

; MLW

, mea

n lo

w w

ater

.

ESC09 9/7/06 3:21 PM Page 318


1 allochthonous – sediments derived from out-side the mangrove and either terrestrial (oftenfluvial) or shallow marine in origin;2 autochthonous – sediments produced in situand which include both organic litter and skeletalmaterial.Sediment accumulation is influenced by a rangeof physical and ecological factors that controltransport, settling and reworking potential.

9.2.4.1 Allochthonous sediments

A significant proportion of the sediment thataccumulates within mangroves derives fromeither terrestrial (mainly fluvially sourced) ornearshore settings. Such sediment may be movedeither as suspended or bedload material (seeChapter 1) and comprises either clastic, carbon-ate or organic material. The relative importanceof sediment sources and of the import/export ofsediment varies considerably between settingsand depends in part on the relative importanceof tidal versus fluvial influence (Fig. 9.11). In themicrotidal Richmond River estuary (Australia)92–99% of the annual suspended sediment loadis derived from fluvial inputs, 90% of whichenters the estuary during short (2 week) sea-sonal flood events (Hossain & Eyre 2002). Only1–2% of estuarine sediment is derived from the continental shelf. In contrast, the Fly Riverestuary (Papua New Guinea) occurs in a meso/macrotidal setting. Despite high fluvial dischargerates (6000 m3 s−1), there is a net inflow of sus-pended sediment from coastal waters whichexceeds fluvial discharge by around 10 times(equivalent to 40 t s−1; Wolanski et al. 1998).The seasonality of river flow is also an import-ant control on the net transport of allochthonoussediment in many fluvially influenced settings.The mesotidal Normanby River estuary in north-ern Australia is, for example, characterized bylong dry seasons (8–10 months) with short wet seasonal flow events (Bryce et al. 1998).Although seasonal flood events are significant,with flood-driven sediment transport estimatedto range from 6000 to 32,000 t per event, this is insufficient to remove all of the annual sandinflux to the estuary (15–30,000 t yr−1 of bedload

sands, and 50,000 t yr−1 of suspended sediment).As a result, the estuary is dominated by net landward flood-tide sediment transport whichaccumulates in the upper estuary.

9.2.4.2 Autochthonous sediments

In addition to externally sourced (fluvial ormarine) sediment, significant amounts of sedi-ment may also derive from in situ sources, themost important of which are organic litter andmangrove-associated skeletal faunas (Fig. 9.11).Most mangroves are characterized by high ratesof primary productivity and, as a result, largeamounts of organic material, in the form of leaflitter along with decaying roots and branches,accumulate in the sediment (Fig. 9.12). Rates ofleaf litter production (a commonly used proxy fororganic matter production; Hogarth 1999) rangefrom 5 to 15 t ha yr−1. Production rates are, how-ever, dependent upon both season and setting.At Darwin Harbour (Australia) Woodroffe et al.(1988) measured the highest litter productionrates (up to 1400 g m−2 yr−1) within tidal creeksettings beneath tall (13 m high) Avicennia sp.trees. In nearby marginal hinterland settings,production of leaf litter near small (< 2 m high)Ceriops sp. was only 300 g m−2 yr−1. At all sites,litter production rates were highest during thewet season.

Leaf litter may be broken down either bymicrobial action or during crab feeding, exportedby tidal or river currents, or incorporated intothe sediment (Hogarth 1999). In a study fromAustralia, Robertson et al. (1992) illustrated the spatial variability that occurs within man-groves in terms of reworking of organic matter.In lower intertidal areas, litterfall averages 556 g m2 yr−1, of which 71% is exported bytidal currents, 28% broken down by crabs andaround 1% lost to microbial action. In contrast,within high intertidal settings, only 33% of the509 g m−2 yr−1 produced is exported by tides,34% broken down by crabs and 33% decom-posed by microbial activity. In settings with ahigh tidal range and/or extensive fluvial activity,significant export of organic material may thusoccur, with the mangrove acting as a major

ESC09 9/7/06 3:21 PM Page 319

320 CHRIS PERRY

source for organic matter (Dittmar et al. 2001).In settings characterized by limited tidal exchangeand by limited fluvial influence, organic matteris largely retained within the environment andrepresents a dominant sediment contributor.Wood detritus breaks down much more slowly,but is facilitated by the activities of wood-boringteredinid molluscs.

Another important input into mangrove sedi-ment derives from mangrove-associated shellyfaunas. Of these the most abundant are crabs,molluscs and foraminifera (Fig. 9.11). Dominantcrab families are the Grapsidae and Ocypodidaeand individual species densities can be in the orderof 60 m−2. Numerous molluscs, including speciesof barnacles, oysters, gastropods and bivalves,also occur and are often dominant skeletal contributors (Plaziat 1995), although there aremarked variations between environments (Plaziat1974). Foraminifera are also abundant, althoughabundance and diversity are strongly influenced

by local hydrodynamics, the seasonality of fluvialinfluence and substrate elevation (Debenay et al.2002). Although the skeletal remains of molluscs,foraminifera and crabs accumulate with man-grove sediments, much of this material is subjectto intense dissolution within acidic porewaters.Although the processes of skeletal modificationand degradation in mangroves remain poorlydocumented, the best preservation is likely inareas where (i) rapid burial in fine-grained (lowpermeability) sediments occurs, or (ii) either sea water or carbonate muds buffer sedimentporewater acidity (Plaziat 1995).

9.2.5 Controls on mangrove sediment transportand accumulation

9.2.5.1 Mangrove sediment transport

The mechanisms and rates of sediment trans-port within and through mangrove systems are

Fig. 9.12 Intertidal sediments accumulating in and around the prop roots of Rhizophora sp. colonies. Note the abundant leaf andseagrass litter accumulating on the substrate. Inhaca Island, Mozambique.

ESC09 9/7/06 3:21 PM Page 320


determined by water circulation patterns. Thesecirculation patterns are, in turn, a reflection ofboth the local hydrology (determined by factorssuch as freshwater runoff and evapotranspira-tion) and tidal regime. Mechanisms of sedimenttransport therefore vary widely depending onthe local climate (which may be highly seasonal)and tidal regime (which varies over both diurnaland monthly spring–neap cycles). In addition,very different mechanisms and rates of sedi-ment transport occur in the tidal creeks and onthe mangrove flats. The tidal creeks often formlong, branched networks and represent the mainconduits of water (and sediment movement). In contrast, the mangrove flats are often wideand heavily vegetated and, because of the highfrictional effects of the vegetation, representsites of sediment trapping and accumulation(see section 9.2.5.2). Ratios of creek to flat areaare in the order of 2–10 in many mangroves(Wolanski et al. 1992) and thus the mangroveflat represents a significant proportion of the tidalprism (especially during the spring-tide phase).Consequently, the processes of sediment trans-port in mangroves exhibit marked temporal andspatial variability.

The tidal cycle represents a particularlyimportant influence on sediment transport bydictating current speeds and the magnitude andfrequency of tidal inundation. Tidal circula-tion is the primary cause of water movementthrough mangrove creeks and there is often a strong asymmetry between ebb- and flood-tide current velocities (see Chapter 1). The ebb-phase is typically shorter and character-ized by stronger current speeds. These may be as much as one-third higher than peak flood currents and exceed rates of 1 m s−1 (Wolanskiet al. 1980). There are also marked differencesbetween the creek networks and the mangroveflats, with current speeds on the flats often lessthan 0.1 m s−1. As a result, significant variationsoccur both in suspended and bedload sedimenttransport (see section 9.2.4.1), as well as in theflux of sediment through tidal creeks and ontothe mangrove flats.

An additional influence on sediment trans-port in the mangrove creeks is the degree of

freshwater versus tidal influence, because thisdictates the extent of saline incursion. Wherethere is strong fluvial outflow, complete flush-ing of saltwater from creeks commonly occursand results in a net outflow of sediment. In areas of restricted or seasonal outflow, sediment isoften retained in upper parts of creek systems(Wolanski et al. 1992). Tidal incursions alsoinfluence the areas over which processes such as secondary circulation, saltwater flocculation,baroclinic circulation and tidal pumping occur.These, in turn, influence rates of sediment trans-port through the tidal creeks (Wolanski et al.1992). Secondary circulation occurs within creeknetworks, where marked vertical differences ineither salinity or suspended sediment loads createdensity gradients (Wolanski 1995; Ridd et al.1998). These are commonly established aroundmeander bends. On the inside of meanders theinterface between density layers may be raisedresulting in differential transport and accumula-tion between the upper and lower parts of thechannel. Under these conditions, finer sedimentsmay accumulate on the upper, inside parts of the bank, and coarser (bedload) material on thechannel floor (Fig. 9.11).

Saline incursions into mangrove creek networksalso influence sediment transport through grainflocculation (fine sediment aggregation). In fresh-water reaches of creeks, fine sediments rarelyflocculate except where organic particles pro-mote grain adhesion and/or suspended sedimentconcentrations exceed 1 g L−1. As a result, sedi-ment settling velocities are influenced by grainsize and density (Wolanski 1995). In saline-influenced creeks (salinity levels > 1), however,saltwater flocculation of fine clay/silt particlesoccurs (Fig. 9.13), with the metallic and organiccoatings on grains promoting grain aggregation.This leads to the formation of large flocs up to200 μm across. Although both clay and calcare-ous flocs have a low cohesive strength, the largergrain sizes increase settling velocities, which maybe up to 100 times those of unflocculated particles(Furakawa & Wolanski 1996). It is likely thatwithout this process most fine sediment wouldtravel through the mangrove as a ‘wash load’(Wolanski 1995).

ESC09 9/7/06 3:21 PM Page 321

322 CHRIS PERRY

The accumulation of flocculated sedimentswithin mangrove creeks is influenced by twoadditional factors: (i) baroclinic circulation and(ii) tidal pumping. Baroclinic circulation occursas a result of the landward movement of denser,saline waters along the bottom of tidal creeks(Fig. 9.13). These currents entrain flocculatedsediments as they settle out and return themupstream towards the limit of saline intrusion(Wolanski 1995). This process is strongest inareas with a pronounced salinity gradient, andresults in most fine sediment remaining in sus-pension near the creek floor. Tidal pumping canoccur in mangrove creeks where there is anasymmetry between the peak flood and ebb-tidal current velocities. The relative strength ofthe flood- or ebb-tide current varies dependingupon tidal regime and the seasonality of themangrove system (see section 9.2.4.1), but insites where tidal currents are stronger during theflood- rather than the ebb-tide phase the result isa net upstream movement of sediment. In com-bination these two processes form a turbiditymaximum zone where fine sediment accumula-tion is concentrated.

As outlined previously not all sediment thatenters mangrove creeks is retained within thecreek network and many mangroves are charac-terized by net sediment export. Sediment exportfrom creek systems will be promoted in systemswhere the frictional effects of high vegetationcover produce a marked surface-water gradient

across the mangrove. This occurs because man-groves reduce flow velocities through the marsh.As a result, flood-tide water levels rise and fallmuch faster in creek and creek margin areas thanon the mangrove flat itself and can create watergradients of up to 1:1000. Where this occurs thetide will be falling at the creek mouth while hightide waters will be ponded on the mangrove flat,producing strong ebb-tide velocities. At RossCreek, northern Australia, threshold velocitiesfor the fine-grained channel substrates are around0.4 m s−1, and flood and ebb-tide current velo-cities are 0.4 and 0.8 m s−1 respectively. As aresult, bedload transport is more prominent dur-ing the ebb-tide phase (Larcombe & Ridd 1995).Under such conditions, sediment will be trappedby vegetation on the mangrove flats duringspring-tide phases, but scoured from creek areasand exported (Wolanski et al. 1992).

9.2.5.2 Mangrove sediment trapping and stabilization

Although tidal and fluvial currents exert animportant influence on sediment transport, sediment accumulation is strongly influenced by mangrove root type and pneumatophoredensity, which modifies current velocities andflow regimes (Woodroffe 1992). As currents passthrough the dense mangrove root networks onthe mangrove flats, the vegetation induces micro-turbulent flow (eddies, jets, stagnation zones),which maintains sediment in suspension. Thismaterial is typically transported landward andsettles out around slack high tide as flow turbu-lence reduces (Furukawa & Wolanski 1996).On the mangrove flats this process is restrictedto periods of high spring tide, and material isprevented from re-entrainment during the ebb-tide phase by vegetation-induced friction. As aresult, tidal currents can act as a ‘pump’, trans-porting sediment landward, so that the highestsedimentation rates occur close to the high tidelimit. This has been demonstrated at MiddleCreek, Australia (Furukawa et al. 1997) wherepreferential sediment trapping occurs in areas ofvegetation-induced flow stagnation and around80% of suspended sediment is trapped in themangroves. This equates to 10–12 kg of sediment

0 10 20 30

162

3

45 Return flow

Turbidity-maximum zone

Salinity

Landward Seaward

Fig. 9.13 Schematic diagram illustrating the processes of grain flocculation and transport associated with barocliniccirculation in a partially stratified estuary: 1, fluvial transport ofunflocculated sediments; 2, saltwater flocculation – large flocssettle out; 3, unflocculated fines remain in suspension; 4, settling of finer grained flocs to the bed; 5, baroclinic circulationtransports settled flocs back upstream; 6, disaggregation ofreworked flocs and re-entrainment in the water column. Theturbidity-maximum zone forms close to the limits of salineintrusion. (Adapted from Wolanski 1995.)

ESC09 9/7/06 3:21 PM Page 322


per metre creek length per spring tide, or a vert-ical accretion rate of around 0.1 cm yr−1. Krausset al. (2003) have documented highest sedimentsettling rates (11.0 mm yr−1) around prop roots,as opposed to pneumatophores (8.3 mm yr−1),although the latter are most effective in terms of sediment retention. Sediment trapping is also aided where bacterial and/or algal matsdevelop on substrate surfaces. Pelletization offine sediment by benthic detritivores will alsobind sediment and limit sediment entrainmentand (re)export.

9.2.5.3 Mangrove sediment reworking

Although small-scale effects of physical sedi-ment reworking and transport influence grainentrainment and transport, biological rework-ing also occurs. Much of this is attributed to theactivities of crabs, which rework sediment dur-ing feeding and burrow construction. Grapsidaecrabs feed primarily on leaf litter, whereasOcypodidae (fiddler crabs) are detritivores andingest sand grains in order to remove organicmaterial. This latter group produce large numbersof faecal and pseudofaecal pellets, with sedi-ment turnover rates of up to 0.5 kg m−2 day−1

(Hogarth 1999). Although this has an importanteffect on retexturing sediments, the generationof extensive, interconnecting burrow networksalso provides a conduit for groundwater flow(Wolanski et al. 1992). Flow velocities in theseburrows may be up to 30 mm s−1 with 1000–10,000 m3 of water flow per tidal cycle persquare kilometre (Ridd 1996).

9.2.6 Variations in mangrove sedimentaccumulation

It is now widely accepted that mangroves tendto follow rather than initiate sediment accumu-lation (Woodroffe 1992), however, once estab-lished the mangroves may enhance sedimentaccumulation by facilitating the trapping of sediment around roots and pneumatophores. Areview of recent literature on short-term verticalaccretion rates by Ellison (1998) suggests thatsediment accretion rates within mangroves are

often less than 0.5 cm yr−1 (with a maximum of around 1 cm yr−1). Sediment accumulationrates are, however, highly variable both withinand between environments, as well as season-ally (Saad et al. 1999). Such differences reflect,in large part, variations in sediment supply and the local processes of sediment transport andreworking. Viewed at a relatively simplistic level,narrow mangrove-fringed shorelines are likelyto exert much less influence on sediment bud-gets than wide mangrove shorelines (Wolanski1995), although few sufficiently detailed budgetstudies have been undertaken to enable detailedunderstanding of the net inputs/outputs of sedi-ment in different mangrove settings.

Different mangrove environments, however,can be identified that have very different sedi-ment substrates and distinct processes of sediment supply and accumulation. The Grijalva–Usumacinta delta, Mexico, for example, forms acomplex network of mangrove-fringed lagoons,channels and interdistributary basins (Thom1967). The delta is developed along a microtidalcoastline and is dominated by fluvial inputs (Fig.9.14a). Fluvial discharge is, however, highly sea-sonal and during the dry season saline wedgesdevelop within the creeks that may extend up to30 km inland. Sediment inputs are dominatedby fluvially derived inorganic sands, silts andclays, and organic material produced in situ fromlitter-fall. In contrast, the mangroves developedin the vicinity of Coral Creek, north-east Austra-lia are associated with an ebb-tide-dominatedestuary, which receives very little fluvial fresh-water or sediment input (Fig. 9.14b; Grindrod& Rhodes 1984; Wolanski et al. 1992). As inthe previous example, the predominant sedimentsubstrate type across the mangrove is an inter-tidal organic mud, derived primarily from in situbreakdown of organic material. Terrigenous sedi-ments are only important in the upper reaches ofcreek networks. Mangroves also develop alongcarbonate-rich shorelines, such as those aroundthe Gulf of St Vincent in southern Australia(Butler et al. 1977). The mangroves develop onmixed carbonate–siliciclastic sediments, althoughmarked variations in sediment content occuracross the mangrove (Fig. 9.14c).

ESC09 9/7/06 3:21 PM Page 323

324 CHRIS PERRY

A

AA LA

R

A ARL

A A A AA

L

R R RA

(a) Deltaic setting

High (seasonal) fluvial inputs

Facies 1 – Distributary channel fill– organic muds + inorganic silts/clays

Facies 2 – Levee deposits– inorganic silts/clays

Facies 3 – Interdistributary basin fills– autochthonous organic sediment

Facies 4 – Mudflats– inorganic fine sand–silt–clay

Limited marine influence

Flood-derived inorganic silts and clays

Periodic breaching transports suspended sediment into adjacent areas

Oxidation reactions lead to iron/calcareous nodule formation

High in situ organic production

Trapping of organic and inorganic sediment around vegetation

Lagoon

(b) Estuarine setting

Facies 1 – Intertidal mudflats– organic muds, silts

Facies 2 – Low tide muds– clay and silt with variable carbonate content

Facies 3A – Upper creek deposits– terrigenous sands

Facies 3B – Middle creek deposits– organic muds, silts

High marine sedimentinputs

Inputs of reworked clastic dune sands

C

B A

R R

C

RR RA

A

Grain flocculation – enhancedby high microbial productivity in nearshore creeks

Extensive trapping of suspended sediment in and around vegetationLimited fluvial

inputs

Facies 3C – Lower creek deposits– shell lags

AA AAA AAA

(c) Open carbonate coast

Facies 1 – Intertidal sands– quartz: carbonate wackestone/packstone

Aeolian sands from surrounding dunes

Extensive binding of sediments by algal mats in upper intertidal zone

Localised precipitation of calcite and aragonite within microbial mats

Sediment trappingaround plant roots

Skeletal carbonateswashed into creeksfrom shallow marine sands

Facies 2 – Nearshore sands– coarse shelly grainstone

Organic inputs(increase in upper intertidal)

Little/no fluvial inputs

High marine sedimentinputs

Fig. 9.14 Schematic diagrams illustrating the sediment facies associated with different mangrove settings. (a) Deltaic setting – basedon the Grijalva-Usumacinta delta in Tabasco and Campeche, Mexico. Mangrove genera: A, Avicennia; R, Rhizophora; B, Bruguiera; C, Ceriops; L, Languncularia. (Based on Thom 1967.) (b) Estuarine setting – based on Coral Creek, north-east Australia. (Based onGrindrod & Rhodes 1984.) (c) Open carbonate-dominated shoreline – based on Spencer Gulf and Gulf of St Vincent, South Australia.(Based on Butler et al. 1977.)

ESC09 9/7/06 3:21 PM Page 324


9.3 PROCESSES AND IMPACTS OF NATURAL

DISTURBANCE EVENTS

Although sediment accumulation within bothcoral reef and mangrove environments is influ-enced by a range of physical and biogenic factors,these sedimentary environments are also sub-ject to natural changes in shoreline morphologyassociated with changing patterns of sedimenttransport and accumulation. They also respondto larger scale and higher magnitude events suchas cyclones and tsunami, and to periodic shiftsin oceanographic conditions such as those linkedto the El Niño–Southern Oscillation. These eventsmay influence the biological components of the respective environments and thus sedimentaccumulation.

9.3.1 Natural dynamics of sediment transport and accumulation

Long-term (Holocene) evolutionary trends alongcoral reef and mangrove-colonized coastlinesreflect complex interactions between eustatic andrelative sea-level change, and a host of externalfactors including, climate, marine environmentalparameters, sediment supply and shoreline geo-morphology. Extensive discussion of these longerterm controls is beyond the scope of this section,but in reef environments a range of evolutionarymodels have been identified, for example, fromstudies of Holocene reef systems (Neumann &Macintyre 1985). In particular, these illustrategrowth response to relative sea-level change.Where reefs initiate at depth, vertical accretionis typical, whereas reefs initiating close to (orreaching) sea-level have tended to prograde laterally. These simple models are often modifiedby local substrate, tectonic or hydrodynamicconditions (Kennedy & Woodroffe 2002).

Mangrove shorelines have also undergoneboth progradation (seaward advance) and trans-gression (shift landward) in response to past sea-level fluctuations (Woodroffe 1992). In general,the post-glacial Holocene sea-level rise resultedin transgression of mangrove shorelines. As sea-levels stabilized during the past 5000 years, areasreceiving sufficient sediment supply subsequently

prograded (Wolanski & Chappell 1996). Thiscontinues in areas that receive abundant allo-chthonous sediment inputs. Examples includethe Ganges–Brahmaputra delta where seawardaccretion rates of 5.5–16 km2 yr−1 have beenestimated (Allison et al. 2003), and the Mekongdelta where some 62,500 km2 of deltaic sedi-ment has accumulated over the past 4500 years(Nguyen et al. 2000).

These rapidly prograding mangrove-colonizedshorelines represent one end-member in a rangeof mangrove settings which exhibit variable shore-line dynamics. This reflects not only spatial andtemporal variability in rates of sediment supply,but also the unconsolidated nature of the sedimentsubstrates and their susceptibility to nearshoredynamics. Alternating phases of progradation andretreat have, for example, been described alongmangrove-fringed coastlines in French Guiana(Froidefrond et al. 1988). This results from long-shore sediment transport, but although indi-vidual sections of this coast alternate betweenphases of accretion and erosion, there remains a net balance in the sediment budget along thewider coastal section (Case Study 9.1).

Along the north-east coast of Australia, mangrove-colonized chenier ridge sequences(see section 9.3.2) have shifted from similarphases of relative shoreline stability (punctuatedby periodic accretion and erosion) to phases ofrapid progradation (Chappell & Grindrod 1984).Prior to 1200 yr BP, coastal evolution was char-acterized by periods of ‘cut-and-recover’. Erosionoccurred during chenier ridge migration, but wasfollowed by renewed small-scale progradation(Fig. 9.15). This period was associated with the development of narrow (< 150 m) mangrovefringes colonizing relatively steep shorelines(slope angle 1:200). Since 1200 yr BP, the coast-line has undergone rapid progradation. Widemangrove fringes have developed along lowerangle shorelines (1:1000) and promote continuedsediment accretion (Fig. 9.15).

In contrast, in north-western Australia there iswidespread net erosion of mangrove-colonizedtidal flats (Semeniuk 1981). Three types of erosioninfluence local and regional intertidal geomor-phology: cliff erosion, sheet erosion and tidal

ESC09 9/7/06 3:21 PM Page 325

326 CHRIS PERRY

Case study 9.1 Cyclical phases of mangrove shoreline accretion and erosion, French Guiana

The nearshore environment along the coast of French Guiana is characterized by the develop-ment of highly mobile mangrove-colonized intertidal mudflats. The sediments that accumulatein these mudflats are derived from the Amazon, which discharges an estimated 1200 × 106 t ofsediment per year. Around 20% of this sediment is transported north-west alongshore by theGuiana Current (Case Fig. 9.1A), which at its peak (January–April) reaches 5 knots and extendsup to 40 km offshore. Sediment is moved both in suspended form (c. 150 × 106 m3 yr−1) andtemporarily stored in the mobile mudbanks that develop along the coast (c. 100 × 106 m3 yr−1).This zone of sediment transport thus creates a mobile nearshore setting, with individual sectionsof the coast alternating between phases of active accretion and erosion.

The distribution of accretionary and erosional sectors is not, however, uniform. Lengths of coastline undergoing active accretion range from 20 to 40 km, are up to 5 km wide at lowtide, and comprise on average an estimated 3 × 1012 m3 of sediment. The mudbanks have amaximum thickness of 10 m and typically exhibit an asymmetric morphology, being steeper on the leeward side. Erosional sectors extend for 8–26 km along the coast and are character-ized by numerous smaller embayments (each up to 500 m wide) giving the eroding sectors a‘sawtooth’ profile. Mudbank migration and associated degradation of the colonizing mangroveforests occurs primarily as a result of ongoing erosion of the windward (up-current) side of the mudbanks and sediment accretion on the lee side (Case Fig. 9.1A). Migration is attributedprimarily to longshore currents and an oblique wave approach. This process, however, may beenhanced in some areas by particularly rapid rates of sediment accretion causing smothering ofmangrove pneumatophores and subsequent mortality.

Magnan BankOrganabo Bank

Iracoubo Bank

Kourou Bank

Cayenne - Mahury Bank

Approuague -Behague Bank

0 km 50

N

Coastal plain

100 km

ATLANTIC OCEAN

Guia

n a Current

FrenchGuiana

Surinam

Guyana

Brazil

Venezuala

Orinoco

Amazon

(a) (b)

Erosion

Accretion

Erosion

Pioneer mangrove

Young mangrove

Mature mangrove

(c)

French Guiana

Submarine section of mudflat

Case Fig. 9.1A (a) Location map showing section ofcoast between the Amazon and Orinoco rivers influencedby the Guiana Current. (b) Detail of inset in (a) showingthe location of intertidal mudbanks along the FrenchGuiana coast. (c) Schematic plan view of an intertidalmudbank showing areas of erosion and accretion, and thedevelopment of mangrove communities across the upperintertidal area. (Adapted from Froidefond et al.1988.)

ESC09 9/7/06 3:21 PM Page 326


Temporal data indicate that sediment mobilization is continual along this coast and thatmudbank migration rates average c. 1.2 km yr−1. Different sectors, however, migrate at differentrates and displacement rates between 1979 and 1984 ranged from 320 m yr−1 (Approuague-Behague bank) to 1220 m yr−1 (Organabo bank). Despite these spatial differences, there appears,to be a net balance along the wider coastal sector in terms of the area of annual mudflat gain(around 60 km2) and erosion (around 58 km2) (Case Fig. 9.1B). One consequence of this highlymobile intertidal setting is to produce a continually changing mangrove environment. Rapidmangrove colonization follows new substrate accretion along the leeward fringes, with theforests continuing to mature until interrupted by later erosion along the windward margins.Interestingly, this transitional community development is mirrored by changes in sedimentaccumulation and sediment diagenetic processes, which change with age of the mudbanks.Within the young (and frequently flooded) mangroves, organic matter in the sediment mainlyderives from algal-mat communities, and diagenetic processes are dominantly suboxic, leadingto rapid degradation of organic matter. As the forests mature, most of the organic matter isderived from higher plant material and diagenetic processes shift to anoxic sulphate-reducingphases, associated with which occurs deposition of pyrite framboids. There is some evidencethat these diagenetic characteristics may be preserved beneath areas of newly accreting mud-banks and thus serve as an indicator of previous erosional and accretionary phases.

Relevant reading

Baltzer, F., Allison, M. & Fromard, F. (2004) Material exchange between the continental shelf and mangrove-fringedcoasts with special reference to the Amazon–Guianas coast. Marine Geology 208, 115–26 (and references therein).

Blasco, F., Saenger, P. & Janodet, E. (1996) Mangroves as indicators of coastal change. Catena 27, 167–78.Debenay, J.P., Guiral, D. & Parra, M. (2002) Ecological factors acting on the microfauna in mangrove swamps.

The case of foraminifera assemblages in French Guiana. Estuarine, Coastal and Shelf Science 55, 509–33.Froidefond, J.M., Pujos, M. & Andre, K. (1988) Migration of mud banks and changing coastline in French

Guiana. Marine Geology 84, 19–30.Marchand, C., Lallier-Vergès, E. & Baltzer, F. (2003) The composition of sedimentary organic matter in relation

to the dynamic features of a mangrove-fringed coast in French Guiana. Estuarine, Coastal and Shelf Science56, 119–30.

Case Fig. 9.1B Areas of land-gain and loss at different sections along the French Guiana coast between 1979 and 1984.(Adapted from Froidefond et al. 1988.)

+ 30

+ 20

+ 10

0

- 10

- 20

- 30

km2

Accretion

Erosion

WNW ESE

App

roua

gue

CayenneKourou

MalmanourySinnamary

IracouboMagnan

Mana

Pte

Iser

e

ESC09 9/7/06 3:21 PM Page 327

328 CHRIS PERRY

creek erosion. Cliff erosion is caused by tidalscour and subsequent mass slumping of channelbanks and occurs primarily along the seawardfringes of mangroves. Rates of retreat average 2 m yr−1, but in places reach 30–50 m yr−1. Sheeterosion occurs due to sediment desiccation andburrowing and results in vertical stripping ofsurface sediments, typically removing a few millimetres at a time (Fig. 9.16). This produceshighly erodible sediment which is removed during spring tides, and rates of vertical lossaverage 1–3 cm yr−1. Tidal creek erosion occursas a result of channel widening and deepening,with lateral erosion of up to 3 m yr−1. Strati-graphical studies indicate that this net erosionalregime has been active over the past 5000 years(Semeniuk 1981).

Smaller scale changes in sediment accumula-tion also occur as a result of local changes influvial sediment discharge or nearshore sedi-ment dynamics, and may result in direct erosionor increased sedimentation. At Portuguese Islandin southern Mozambique, changes in coastlineconfiguration have caused both direct erosion of mangrove substrates and progressive tidalrestriction, leading to a 75% reduction in man-

grove extent between 1958 and 1989 (Hatton &Couto 1992). Conversely, sediment mobilizationalong mangrove channels in Brunei, Malaysiahas caused channel damming and increasedmangrove flooding, leading to localized mortal-ity of mangrove species and intertidal fauna(Choy & Booth 1994). A common impact ofchanging sediment dynamics is modified sedi-ment accumulation rates. Although many man-grove species can tolerate accretion rates of5–10 mm yr−1, higher rates may result in man-grove mortality due to root smothering (Ellison1998). Where this occurs, intertidal substratesmay become unstable and prone to reworking.Mangrove shorelines are often, therefore, highlydynamic in areas of either high allochthonoussediment input, or where intertidal sedimentsare subject to frequent mobilization.

In reef environments, rapid geomorphologicalchange, in terms of the position of the reef bodyand its shoreline geometry, are often less evident.This probably reflects the more rigid structure ofmany framework-dominated reefs, which requiresignificantly elevated energy regimes, such asthose associated with cyclone activity, to facilit-ate sedimentary responses (although under such

Shell beds

Cyclone

Old chenierridge

New chenierridge

T1 - Progradation

T2 - Cyclone-driven shorelineerosion and chenier ridge emplacement

T3 - Continued rapid progradation

Cut and recover mode Rapid prograding mode

Cyclone

T1 - Rapid progradation

T2 - Cyclone-driven emplacement of chenier ridge, but little/no shoreline erosion

T3 - Progradation restores shoreline

Mangrove damage

Fig. 9.15 Contrasting models of shoreline evolution at Princess Charlotte Bay, northern Australia. (Adapted from Chappell & Grindrod 1984.)

ESC09 9/7/06 3:21 PM Page 328


conditions major sedimentological changes canoccur – see section 9.3.2). Coral communitiesthat occur within some nearshore, turbid settingsmay, however, be susceptible to episodic sedi-ment mobilization. At Inhaca Island in southernMozambique, coral communities (but not frame-work reefs) are patchily developed along themargins of shallow intertidal channels (Perry2003). These communities undergo episodicmortality due to smothering and burial of coralcolonies associated with longshore sedimenttransport. In turn, new areas of bedrock andcoral rubble may be exposed and subsequentlycolonized (Perry 2005). The natural character-istics of coastal sediment dynamics thus createan ephemeral suite of coral communities.

9.3.2 Physical disturbance (cyclones and tsunami)

Although background or seasonal fluctuationsin sediment transport processes exert a clearinfluence upon reef and mangrove systems,

significant changes also occur in response tolow-frequency, high-magnitude events such ascyclones and tsunamis. Tropical cyclones (alsotermed hurricanes and typhoons, dependingupon geographical location) are characterized bycyclonic surface winds formed around centres of low pressure (McGregor & Nieuwolt 1998).These weather systems may be up to 800 km indiameter and are characterized by strong winds(often > 30 m s−1). Impacts on nearshore sedi-ments, and on reef and mangrove communities,result from strong wind-driven currents, highwave heights (5–15 m), and elevated coastalsea-levels.

Physical disturbances to both coral reef andmangrove communities are well documentedconsequences of cyclones. In coral reefs, wavedamage may result in widespread breakage, top-pling and abrasion of shallow water (especiallybranching) corals (Rogers 1993), while high windspeeds and storm-wave surges into intertidal areascause mangrove uprooting and damage (McCoy

Fig. 9.16 Severe sheet erosion leading to removal of the upper layers of mangrove sediment and the exposure of root networks.Gordon Creek, Australia. (Photograph courtesy of Piers Larcombe.)

ESC09 9/7/06 3:21 PM Page 329

330 CHRIS PERRY

et al. 1996). The scale and extent of disturbanceare often highly variable even over short spatialscales, but in high-impact areas reduced livecoral or mangrove cover may be an immedi-ate consequence of disturbance. Disturbancemay be exacerbated in reef environments by thesubsequent effects of cyclone-related rainfall,which may induce localized coral bleaching (VanWoesik et al. 1995), and in mangrove settings,increased rainfall may reduce soil salinity levelsor cause erosion due to surface runoff.

From a sedimentological perspective, cyclonesexert a direct influence on sediment transport,and, over geological time-scales, should beregarded as an important control in tropicalsedimentary environments. A detailed review of the geological effects of cyclones within reefenvironments is provided by Scoffin (1993), whoidentified a range of erosional and depositionalprocesses. These include both on- and offshoretransport of sediment and rubble, much ofwhich is derived from physical breakdown ofthe coral communities. Onshore transport can

result in the development of rubble cays orridges (Fig. 9.17; Hayne & Chappell 2001) andsediment deposition in the lee of these features(Fig. 9.18), whereas off-reef transport produceslocal talus deposits. In shallow water, storm-generated rubble can be a major constituent ofreef facies (Blanchon et al. 1997). Storm surgesmay also result in significant sediment mobiliza-tion in shallow water, leading to beach, coralcay and subtidal erosion.

Studies in St Croix provide an insight into thevolumes of sediment that may be mobilized duringhurricanes. At Salt River Canyon, an estimated2 × 106 kg of sediment were flushed from the reefsduring the passage of Hurricane Hugo in 1989,and at nearby Cane Bay an estimated 336,000 kgof sediment were flushed offshore from a singlereef channel (Hubbard 1992). Recent work onthe Great Barrier Reef (GBR) shelf demonstratesthe extent to which cyclones also influence sedi-mentation across much wider carbonate shelfsystems (Larcombe & Carter 2004; see alsoChapter 10). Cyclones, which typically approach

Fig. 9.17 Storm-generated coral rubble ridges. A series of ridges, up to 3 m high, occur at these sites, recording evidence of successivestorm events. Triangulos, Campeche Bank, Gulf of Mexico.

ESC09 9/7/06 3:21 PM Page 330


the Queensland coast from the north-east, drivealong-shelf currents that transport sedimentnorthwards along the shelf, resulting in markedcross-shelf variations in Holocene sedimentaccumulation (Case Study 9.2).

Along mangrove-colonized shorelines, the mainsedimentological impacts of cyclonic disturbanceresult from shoreline erosion caused by stormsurges. In Florida, some 15 m of shoreline ero-sion occurred following Hurricane Andrew in

LagoonReef crest/flat

Reef front

Deeper water corals dislodged bymaterial flushed down the reef front

Fragmentation of shallow water branched corals - rubble moved both landward and offshore

Boulder emplacementon reef flats

Talus deposits ofcoral rubble at base of reef front

Off-reef transport offine sediments - settle in deeper water

Accumulation of overwashsand/rubble lobes in the lee of reef flat

Scouring of shoreline sands

Reworking/transportof shallow lagoon sands

Deposition of successiveridges composed of coral rubble removed from reef front and reef crest sites

Overturning of coral bommies

Fig. 9.18 Schematic diagram illustrating the main erosional (bold italics) and depositional (italics) processes associated with cycloneimpacts on coral reefs.

Case study 9.2 Cyclone-controlled sediment distribution on the Great Barrier Reef shelf (the ‘cyclone pump’ model)

The Great Barrier Reef (GBR) shelf is the largest modern tropical mixed carbonate–siliciclasticshelf system, extending around 2000 km along the eastern and north-eastern margins ofAustralia, and varying in width from around 200 km in the south to 50 km in the north. It ischaracterized by extensive reef development along its seaward margin, but also by sedimentdeposition on the shelf. In the central regions, the shelf can be partitioned into three distinctshore-parallel sediment zones. The inner shelf (0–22 m depth) is dominated by terrigenoussands and muds (5–10 m thick) and includes the development of intermittent detrital coralbuild-ups (Smithers & Larcombe 2003). Under fair-weather conditions, sediment transport isdominated by the northwards, along-shelf current-driven movement of sediments resuspendedby wave action in nearshore areas, and bedload transport occurs only in the shallow (< 5 m)shoreface zone. The middle-shelf (22–44 m depth) is a zone of sediment starvation and is characterized by only a thin veneer (< 1 m thick) of Holocene shelly, muddy sands. This zonereceives little or no terrigenous sediment and there is limited sediment transport under fair-weather conditions. The outer-shelf (40–80 m depth) is dominated by a tract of reef develop-ment and by associated carbonate-dominated detrital sediments. In fair weather little or noterrigenous sediment reaches the outer reefs (Case Fig. 9.2a).

Cyclones are a common disturbance on the central GBR (two to three cyclones occur each year at latitude 20°S) and therefore represent a major control on shelf sedimentation.Cyclones influence sediment input to the GBR by:

ESC09 9/7/06 3:21 PM Page 331

332 CHRIS PERRY

1 causing the breakage of reef framework on the outer shelf, which contributes carbonate sandand gravel on the leeward flanks of individual reefs;2 generating floods in rivers and discharge of terrigenous sediment to the inner shelf;3 eroding the sea-floor of parts of the inner and middle-shelf zones (Case Fig. 9.2a & b).Cyclones also, however, exert a major control on the mobilization and transport of sedi-ment on the shelf. Under cyclonic conditions, fast wind-driven currents (> 60 cm s−1) have beenrecorded and, in the case of Cyclone Joy (1990), these formed a persistent flow of 9 days dura-tion. Under such conditions, extensive along-shelf transport of sand and gravel occurs, and thecyclones act as a ‘pump’, transporting large volumes of sediment northwards along the shelf(Larcombe & Carter 2004).

Sedimentary evidence for such transport occurs across much of the inner and middle shelf, in the form of large (up to 2 m high) subaqueous sand and gravel dunes, sediment ribbons andlarge dune fields. Extensive erosion of the sea-floor also occurs across the middle shelf, which incombination with high rates of cyclone-driven sediment transport may explain the lack of reefdevelopment on the middle shelf throughout the Quaternary. The cyclones also redistributelarge amounts of the suspended terrigenous sediment that is discharged onto the shelf duringstorm events, driving this sediment along and onshore, and thus contributing to the develop-ment of the terrigenous sediment-dominated inner shelf. In this way, cyclones act both as a control on cross-shelf sediment accumulation and along-shelf sediment transport.

Relevant reading

Gagan, M.K., Chivas, A.R. & Herczeg, A.L. (1990) Shelf-wide erosion, deposition, and suspended sedimenttransport during Cyclone Winifred, central Great Barrier Reef, Australia. Journal of Sedimentary Petrology60, 456–70.

Larcombe, P. & Carter, R.M. (2004) Cyclone pumping, sediment partioning and the development of the GreatBarrier Reef shelf system: a review. Quaternary Science Reviews 23, 107–35.

Smithers, S. & Larcombe, P. (2003) Late Holocene initiation and growth of a nearshore turbid-zone coral reef:Paluma Shoals, central Great Barrier Reef, Australia. Coral Reefs 22, 499–505.

Townsville

Cyclone impacts:(1) Breakage of reef framework and mobilization of reef sediment (2) Cyclone 'pumping' of sediment along and from the middle shelf(3) Fluvial sediment inputs onto the inner shelf

A

A'A A'

Coral reefs

Inner shelf - terrigenous sands/muds

Middle shelf - shelly muddy sands

Outer shelf - carbonate sands

(a) (b)

(2)

(2)

(2)

(3)

(3)

(1)

(1)

0 5 10 15 20 25 30 35Distance across shelf (km)

(3)

(1)(2)

(2)(2)

01020304050

De

pth

(m

)

Uppercontinentalslope

Case Fig. 9.2 Simplified plan (a) and cross-sectional (b)views showing the effects of cyclones on sediment generationand dispersal across the central Great Barrier Reef. (Adaptedfrom Larcombe & Carter 2004.)

ESC09 9/7/06 3:21 PM Page 332


1992 (Swiadeck 1997), although this is likely tohave been mitigated by the presence of nearbyprotective reef flats, and by the relatively restrictedfetch of the approaching waves. Well-documenteddepositional features of cyclone-influenced tropi-cal shorelines are chenier ridges. These comprisesands and shells winnowed from adjacent mud-flats and transported through the mangroves by waves during cyclone-related storm surges.Individual cyclones may result either in theemplacement of chenier ridges or the migrationof ridges across chenier plains (Chappell &Grindrod 1984; Woodroffe & Grime 1999).Cyclone-related sedimentation, driven by storm-surges, can also result in widespread mangrovemortality due to root burial (Ellison 1998).

High-energy conditions are not confined tocyclones, with significant physical disturbance alsoassociated with tsunami. The long-wavelengthwaves associated with tsunami are generated byearthquakes, volcanic eruptions or submarineslides and can move at speeds in excess of 700 kmh−1 in the open ocean. As water depth decreasesnearshore wave height increases dramatically,resulting in significant nearshore damage andsediment and rubble transport (Solomon & Forbes1999). The high-magnitude (Richter scale 9.0)earthquake-generated Indian Ocean tsunami of 26 December 2004 provided dramatic evid-ence of both the major geomorphological andsedimentary impacts that these events can have(Fig. 9.19), as well as the high human and socio-economic cost. The impact of this event on coralreef communities appears, as is commonly thecase with cyclone damage, to have been spatiallyvery variable, with major coral damage reportedat some sites, whereas other areas survived un-scathed. At the time of writing a number of post-event geomorphological and geological surveyswere being conducted, and these may provide auseful insight into the role (or not) of reefs andmangroves in affording protection to affectedshorelines. Common sedimentological indicatorsof past tsunami events (tsunamites) include theoccurrence of large limestone megaclasts on reefflats, along with the occurrence of allochthon-ous reworked sands and shells (Noormets et al.2002; van den Bergh et al. 2003).

9.3.3 Large-scale oceanographic changes

In addition to the physical disturbance associatedwith high-energy events, major changes in near-shore environmental conditions periodically occurdue to shifts in oceanographic conditions. Themost common example of this is associated withthe El Niño–Southern Oscillation (El Niño) whichcauses major changes in sea-surface temperaturesand climate in the Pacific Basin (McGregor &Nieuwolt 1998). During normal (non-El Niño)years, warm water in the equatorial Pacific isdriven westwards by prevailing trade winds,resulting in higher and warmer sea-surface con-ditions in the west. The westward flow is com-pensated for in the eastern Pacific (i.e. the westernseaboard of central America) by upwelling ofcool, nutrient-rich waters. These waters limitextensive reef development (see section 9.1.3).During El Niño years, however, the trade windsweaken and warm water sloughs back eastwards,restricting upwelling and bringing warm watersinto the nearshore areas of the eastern Pacific.This further influences the coral communities,via coral bleaching and mortality as sea-surfacetemperatures increase. There are, however, alsosignificant implications for reef carbonate budgets.Studies in Panama (Eakin 1996) documentedmajor reductions in reef carbonate productionfollowing the 1982–83 El Niño. Before 1982, reefsaround Uva Island were characterized by netcarbonate production rates of 0.34 kg m2 yr−1,but after 1983 a 50% reduction in coral cover,along with continued high rates of bioerosion, ledto net reef erosion (average of −0.19 kg m2 yr−1).Sea-level fluctuations of up to 0.5 m between ElNiño years can also result in significant remobi-lization of nearshore sediments and shorelineerosion (Solomon & Forbes 1999), as changes inclimatic conditions are marked by significantlyincreased rainfall and fluvial runoff.


DISTURBANCE

Although most reef and mangrove environmentsare subject to natural physical reworking and

ESC09 9/7/06 3:21 PM Page 333

334 CHRIS PERRY

(b)

(a)

Fig. 9.19 Quickbird satellite imagery showing the effects of the 26 December 2004 tsunami in the vicinity of Gleebruk, south of BandaAceh, Indonesia. (a) Pre-tsunami image taken on 12 April 2004. (b) Post-tsunami image taken on 2 January 2005. As well as evidence ofmajor damage to fields and villages, there is substantial evidence of major shoreline change, with erosion of beaches (boxed areas) andmajor changes to river and channel mouths (arrowed). (Imagery courtesy of DigitalGlobe (http://www.digitalglobe.com).)

ESC09 9/7/06 3:21 PM Page 334


changes in sediment flux, some significant changesare also linked to anthropogenic disturbance.Many such disturbances may have an immediateimpact on the reef or mangrove community, butsome may also have longer term sedimentologicalimpacts. In the context of coral reefs, this relatesto changes in rates and patterns of carbonateproduction, in the case of mangroves to the sub-strate destabilization that can follow communitydegradation, and in both environments to pro-gressive sediment contamination. A key issue interms of disturbance relates to the need to placeobserved environmental or community changesin the context of natural or background commun-ity dynamics, i.e. change should not necessarilybe regarded as a consequence of anthropogenicdisturbance. This section is not intended as acomprehensive review of anthropogenic influ-ences but outlines, via selected examples, someof the main impacts (and uncertainties) of humandisturbance in relation to reef and mangrovesedimentology. Useful reviews of potential anthro-pogenic disturbances to mangroves and coralreefs are provided, respectively, by Ellison &Farnsworth (1996) and Grigg & Dollar (1990).

9.4.1 Impacts of coastline modification andresource extraction

Modification of the coastal fringe as a result of land-use change, construction or resourceextraction can exert a major influence on sedi-ment dynamics and coastline stability (see alsoChapters 7 & 8). Change may occur eitherdirectly due to sediment removal, leading tochanges in local sediment budgets, or indirectlyas a result of seawall or causeway construction,which may modify local current dynamics andrestrict longshore sediment transport (Preu1989). In reef environments, progressive habitatdegradation resulting from coral mining, sandextraction or channel blasting may modify localcurrent pathways, and in mangrove settings,land conversion for agriculture may removeareas from tidal influence and thus alter the area over which tidal influence is exerted. Suchchanges can lead to modified patterns of sedi-ment transport, erosion and deposition.

In south west Sri Lanka a range of human-related activities have promoted rapid coastalretreat (average retreat 1.1 m yr−1; Preu 1989).Activities include large-scale fluvial sand extrac-tion, which prevents sediment (normally sup-plied to the beaches) from reaching the coast. In addition, reef degradation resulting fromchannel blasting and coral mining has led tomodified nearshore current dynamics and locallyhigh (up to 4 m yr−1) rates of shoreline retreat.Shoreline erosion rates of 0.5–1 m yr−1 havealso been reported from the Marshall Islands,where aggregate extraction has reduced sedimentsupply to the lagoon rim, and causeway andchannel constructions have interrupted sedimenttransport (Xue 2001).

Changes in land-use and the transition ofmangrove land for agricultural or developmentpurposes also have an impact on nearshore sedi-ment dynamics. In particular, loss of mangroveswamp has an impact on the spatial extent of tidalinfluence. This may reduce the tidal prism leadingto modified current dynamics, reduced ebb-tideflow and sediment export (see section 9.2.5.1),and thus channel siltation (Wolanski et al. 1992).Land modification for the purposes of shrimp orrice farming also has longer term impacts beyondthe immediate loss of mangrove substrate. Thesediments in such areas typically become highlysaline and acidic, and this often inhibits naturalrecolonization. This in turn leaves areas of bare, degraded substrate which are more proneto erosion from runoff and waves (Ellison &Farnsworth 1996).

9.4.2 Impacts of increased sediment flux

Although high rates of sediment discharge can beregarded as natural inhibitors or influences on reefand mangrove development (see section 9.1.3),increased or modified rates of sediment influxare a widely cited example of anthropogenic dis-turbance (Grigg & Dollar 1990). This has beenattributed to poor land management practices(which may increase soil erosion and runoff) andto marine dredging and dumping. Two key factorsneed, however, to be highlighted in relation tosediment inputs into nearshore environments:

ESC09 9/7/06 3:21 PM Page 335

336 CHRIS PERRY

sedimentation and turbidity. The former is ameasure of downward sediment flux, the latter ameasure of suspended sediment load (Te 1997).The two are not, however, directly correlatedbecause high turbidity conditions can occur inareas with relatively low sedimentation rateswhere wave-driven resuspension rates are high(Thomas et al. 2003). Both factors, however,have a potential impact on coral communities;increased sedimentation leads to smothering andburial, and to an increase in polyp stress, whereasturbidity acts to inhibit light penetration andthus restrict photosynthesis. As a result, sedi-ment influx has been widely perceived as a keythreat to coral communities (see Rogers (1990)and references therein).

Recent research, however, has highlightednumerous sites where coral communities (andreefs) not only develop but also persist, underconditions of high turbidity and periodicallyhigh sediment flux (Woolfe & Larcombe 1998;Perry 2003; Smithers & Larcombe 2003). These‘reefs’ are typically characterized by reducedframework development (although not neces-sarily reduced coral species diversity) and, as aresult, are somewhat distinct from the coral reefsthat can develop in clear-water settings. Theyrepresent, however, locally important sites ofcoral community development and bring intoquestion widespread assumptions about the nega-tive effects of sedimentation. Coral reefs occur,for example, at a number of sites along theinshore regions of the Great Barrier Reef, despitehigh rates of terrigenous sediment influx (Woolfe& Larcombe 1998; Smithers & Larcombe 2003).Furthermore, several studies present evidencesuggesting that coral communities have con-tinued to develop under conditions of at leastperiodically high turbidity and terrigenous sedi-ment input for as much as the past 5000 years(Johnson & Risk 1987; Perry 2005).

Although local increases in the amount ofsediment reaching coral reefs (especially in areas where past inputs have been limited) areundoubtedly likely to have a negative impact on corals and to modify reef community struc-ture, too little is known about the occurrence of coral communities under natural conditions

of high turbidity and elevated sedimentationrates to generalize about sediment input, coralresponse and reef occurrence. Such generaliza-tions are further complicated by the vastly dif-fering regimes of sediment input and sedimentcomposition that characterize different coastalsites. Sediment inputs vary between sites depend-ing upon the frequency (seasonal, episodic),longevity and volumes of terrigenous sedimentinput, and sedimentation rates and turbidityregimes vary depending upon grain size, degreeof sediment resuspension and tidal range. Theselocal variations will influence coral speciesresponse to sedimentation in terms of sedimentrejection mechanisms and growth strategies.Where sediments do influence coral growth, theyare likely to be significant from a sedimento-logical perspective because rates and patterns ofreef carbonate production are directly influencedby the composition of the reef community.Studies in Indonesia have, for example, docu-mented net erosion on a range of reefs subject tohigh terrigenous sediment and nutrient inputs(see also section 9.4.2) and this has been attri-buted both to reduced coral cover and increasedrates of bioerosion (Edinger et al. 2000).

Despite their association with fine, organic-rich sediments, mangroves are also potentiallysusceptible to increased sedimentation. Ellison(1998) reviews numerous examples of increasedsediment influx relating to dredge spoil dump-ing, construction-related sedimentation, miningand catchment deforestation. The primary im-pacts on mangrove communities relate to burialof aerial roots at rates sufficient to restrict soil-gas exchange. This can cause mangrove mortal-ity and thus loss of mangrove cover, and erosionin areas subject to fluvial or marine reworking.Sedimentation rates in excess of 1 cm yr−1 arelikely to be detrimental, although different speciesof mangroves exhibit different tolerances toburial (Thampanya et al. 2002).

9.4.3 Impacts of increased nutrient input

The primary source of nutrients into shallowmarine and intertidal environments is fromanthropogenic sources (e.g. sewage effluent,

ESC09 9/7/06 3:21 PM Page 336


agricultural runoff), the most important nutri-ents being nitrogen and phosphorus. In coralreefs, ecological impacts of excess nutrient inputsinclude suppression of coral growth rates andreproductive capacity (Tomascik & Sander 1985,1987). In many cases, nutrient inputs are coupledwith increased sediment inputs and this com-plicates isolation of causal mechanisms of dis-turbance. A possible consequence of nutrification,however, is to increase the competitive advant-age of opportunist algal species over corals. Thisis significant from a sedimentological perspec-tive in relation to changing the composition andabundance of the carbonate-producing com-munity. The potential impacts of such stressesare highlighted in studies from the Florida Keys,where water eutrophication has been cited as amajor cause of temporal change in carbonatesedimentation (Case Study 9.3).

Available data indicate that the effects ofnutrification may be less significant in mangrove

settings, although this may vary in differentmangrove environments. Mangroves commonlygrow in waterlogged, anaerobic soils, so thatanaerobic reactions exert a major influence onphosphorus and nitrogen chemistry. These reac-tions are strongly influenced by redox potential,which varies depending on the degree of tidalinundation, sediment porosity and organic mattercontent (Clough et al. 1983). Phosphorus occursprimarily in organic form and, in most man-grove sediments, low redox potentials lead torelease of phosphate. Therefore, in systems thatreceive additional phosphate associated withnutrient inputs, the ability to immobilize phos-phorus may be limited. Nitrogen occurs mainlyin organic form (primarily ammonium), and any nitrate is rapidly converted by anaerobicbacteria (dentrification) to gaseous nitrogen ornitrous oxide. Most of this occurs in interstitialwaters within mangrove sediments and is highlysusceptible to leaching by rainwater or during

Case study 9.3 Decadal-scale changes in carbonate sediment compositions across the Florida reef tract, USA

Coral reef communities along the Florida reef tract have been subject to a range of both naturaland anthropogenic-related disturbance events over the past 40 years. These have includedphysical damage from hurricanes and extreme cold-water events, as well as the effects of coral disease (White Band Disease, Black Band Disease) and mortality of the key herbivorousechinoid Diadema antillarum. In addition, there is evidence that Florida reef-tract waters arebecoming increasingly eutrophic as a result of nutrient inputs and infiltration from ground andsurface waters (Szmant & Forrester 1996). The consequence of this has been a general declinein reef ‘health’ as evidenced by widespread reductions in live coral cover and consequentincreases in algal biomass.

Comparisons of sedimentary data collected from sites along the Florida reef tract (Case Fig. 9.3) in 1963 with that collected in 1989 indicate differences in the abundance of biogenicsediment constituents that may reflect this progressive decline in coral reef ‘health’ (Lidz &Hallock 2000). On a reef-wide scale, the major changes in composition have been a doubling inthe relative abundance of molluscan grains, and a tripling in abundance of coral fragments.Marked variations occur, however, in different parts of the Keys, reflecting spatial variations inreef vitality. For example, the proportion of coral grains in Upper Keys sediments (a relativelyhealthy section of the reef) is reduced compared with the large increases that have occurred in the Middle and Lower Keys (where coral communities are in decline) (Case Fig. 9.3b). Thereis little evidence to suggest that these changes reflect periodic storm/hurricane reworking of sediment, but rather are attributed to increased rates of dead coral substrate erosion by internal

ESC09 9/7/06 3:21 PM Page 337

338 CHRIS PERRY

bioeroders (borers). This has, in large part, been linked to increased nutrient levels in nearshorewaters as a result of increased sewage discharge. This stimulates phytoplankton productionand in turn rates of bioerosion by suspension-feeding boring organisms.

In a more localized study around Key Largo, Cockey et al. (1996) also identified increasingnutrient levels in nearshore waters as a likely cause of temporal changes in the composition offoraminiferal assemblages within reef-related sediments. Samples from 1959 to 1961 indicateda predominance of larger, symbiont-bearing Soritidae foraminifera, which comprised some50–80% of the species identified. Samples recovered from the same sites in 1991–92, however,showed an overwhelming dominance of smaller, heterotrophic Miliolidae and Rotaliidae.These studies illustrate the close linkages that exist between the reef community and productionof carbonate sediment. Major changes in the abundance of either carbonate producers ordegraders can result in detectable shifts in the abundance of carbonate sediment constituentsover decadal time-scales. These studies illustrate the potential of sedimentary data as geo-indicators of environmental change.

Relevant reading

Cockey, E., Hallock, P. & Lidz, B.H. (1996) Decadal-scale changes in benthic foraminiferal assemblages off KeyLargo, Florida. Coral Reefs 15, 237–48.

Lidz, B.H. & Hallock, P. (2000) Sedimentary petrology of a declining reef ecosystem, Florida reef tract (U.S.A).Journal of Coastal Research 16, 675–97.

Perry, C.T. (1996) The rapid response of reef sediments to changes in community structure: implications fortime-averaging and sediment accumulation. Journal of Sedimentary Research 66, 459–67.

Szmant, A.M. & Forrester, A. (1996) Water column and sediment nitrogen and phosphorous distribution patterns in the Florida Keys, USA. Coral Reefs 15, 21–42.

Florida Bay

BiscayneBay

Hawk Channel

South Florida

Straits

of F

lorid

a

Lower Keys Middle Keys

Upper Keys

20 km

N

(a)

−36

mGulf ofMexico

Area shown in detail in (b)

Long Key

Key Largo

20 km

Long Key

Key Largo

1963

1989

Dominant grains

HalimedaMolluscCoral

(b)

Case Fig. 9.3 (a) Simplified map showing the main geomorphological features of South Florida. (b) Sediment data collected in1963 and 1989 from the Middle Keys area showing greater abundance of coral within the sediments in 1989. (Adapted from Lidz& Hallock 2000.)

ESC09 9/7/06 3:21 PM Page 338


tidal inundation. Such processes may be enhancedby the high levels of particulate organic matterthat are often associated with sewage or agri-cultural discharges and may speed up anaerobicreactions, increase rates of dentrification andhave an impact on mangrove growth. Althoughmangroves may actually benefit from increasednutrient inputs (and exhibit increased produc-tivity), nutrient enrichment of soils may have amore negative impact by altering soil chemistryand increasing the anaerobic nature of the sedi-ment (Field 1998).

9.4.4 Impacts of contaminant input (oil,chemicals, metals)

Common contaminant inputs into nearshore andintertidal environments include a range of heavymetals, as well as organochlorine pesticides andoil residues. These may be derived from indus-trial, urban and agricultural discharges, or fromcontaminant dumping and refuse sites. Contam-inant transfer may occur either in suspended ordissolved form, with widespread distribution in tidally influenced settings being enhanced bydiffusive flow mechanisms (Wolanski et al. 1992).Transfer through sediments can occur due toseepage and runoff (Clark 1998). Contaminantaccumulation is aided, particularly in mangroves,by the fine-grained nature and high organic content of the sediments (Harbison 1986; seealso Chapter 7). Mangrove sediments act bothas physical traps for fine particulate and trans-ported contaminants, and as chemical traps wheremetal sulphides precipitate from solution. Suchprecipitation is driven by a range of bacterialsulphate reduction reactions in the anaerobicsediments. Although there is the potential forprecipitated metals to be remobilized into sur-ficial sediments and overlying waters (Harbison1986), precipitation of metal sulphide is environ-mentally significant as a mechanism of immob-ilizing metal contaminants.

Gradients of metal contamination within sedi-ments are common through mangrove swamps(Soto-Jiménez & Páez-Osuna 2001), with thehighest metal concentrations tending to occur inareas rich in fine particulate and organic matter.

The extent to which metal enrichment of sedi-ments has a direct impact on the mangrove com-munity is not clear. Some studies report uptake ofmetals within mangrove roots, often to levels thatexceed surrounding sediments (e.g. MacFarlaneet al. 2003) but these do not appear to have animpact on mangrove growth. In contrast, uptakewithin leaves appears low (Machado et al. 2002),but in the context of metal export from man-groves this is significant given the high amountsof leaf detritus that are exported from somemangrove systems (see section 9.2.4.2).

Reef sediments also have potential to accumu-late metal and chemical contaminants, and thereis evidence of marked spatial variations relatedto sediment grain size. On the inshore GreatBarrier Reef higher metal concentrations occur,for example, within fine-grained, clay-rich har-bour sediments compared with coarse-grainedcarbonate–quartz sands (Esslemont 2000). Atthese sites, remobilization and transport of fine-grained, metal-rich sediments onto adjacentcoral reef areas has occurred periodically due to harbour dredging. Incorporation of metalsinto coral skeletons may occur via (i) absorptionof dissolved metals by coral tissue or duringfeeding, (ii) particulate trapping in cavities or(iii) direct deposition on coral skeletons throughpolyp damage (Fallon et al. 2002). Such uptakehas documented ecological impacts upon coral fertilization and coral larvae success (Reichett-Brushett & Harrison 1999), although long-termeffects on skeletal organisms remain poorly documented.

In extreme cases, inputs of contaminants alsohave potential to cause changes in carbonatesediment assemblages via progressive sedimentdilution. In north Jamaica, bauxite dust from a loading terminal built in the mid-1960s hasaccumulated within sediments in Discovery Bay,a semi-restricted coastal embayment fronted by fringing reefs and previously dominated bycarbonate sediments. Recent studies (Perry &Taylor 2004) have shown that bauxite now com-prises upwards of 35% of the sediment in themost heavily impacted areas and some 20–25%of the sediment across much of the southern/central area of the bay (Fig. 9.20).

ESC09 9/7/06 3:21 PM Page 339

340 CHRIS PERRY

Significant ecological impacts on mangrovesand corals also can be attributed to oil con-tamination. In large part, impacts relate to thebuoyant nature of spilled oil, which is thus easily transported into intertidal settings bywinds and currents. In mangroves, ecologicalimpacts relate primarily to smothering of aerialroots and pneumatophores, leading to oxygendeficiency and rapid mortality of mangrovetrees (Levings & Garrity 1994). There also maybe impacts on mangrove propagule vitality(Duke & Watkinson 2002). In reef settings, oilmay reduce coral growth and modify reproduc-tive rates (Guzmán et al. 1994). The incorpora-tion of oil residues into shallow marine andintertidal sediments is strongly site-dependentand is influenced by the rate of hydrocarbonweathering (Munoz et al. 1997) and the tidalregime, which influences the spatial extent of oildistribution (Lewis 1983). It is clear, however,that mangrove sediments have the potential to

act as temporary sinks for oil residues. This isattributed in part to the anaerobic nature of thesediments, which inhibits microbial degradationof hydrocarbons, and to the rapid incorporationof oil into the sediments via burrow networks.Trapping is also aided by the ability of hydro-carbon residues to bond tightly to fine-grainedsuspended sediments (Ke et al. 2003). The long-term persistence of oil residues in mangrove systems is discussed by Burns et al. (1993), who also illustrate the potential for periodic orpulsed phases of oil residue release as sedimentsare leached during runoff or storm-wave-relatederosion (Case Study 9.4).

9.4.5 Multiple-cause disturbances and changingpatterns of sedimentation

Particularly in the case of coral reefs, it is often difficult to isolate the individual factorsthat induce community decline and shifts in

Shi

p ch

anne

l50 m

40 m

5 m

30 m20 m

10 m5 m

20 m10 m

5 m

JAMAICA

DiscoveryBay

N

500 m Bauxite loadingterminal

Reef crest(a) > 95%

90.0–94.9%85.0–89.9%80.0–84.9%75.0–79.9%<75.0%

Carbonate content(b)

Fe vs carbonate

40

50

60

70

80

90

100

0 5000 10000 15000Fe (µg g−1)

Car

bona

te (

%) r2 = 0.811

(c)

Fig. 9.20 (a) Location of Discovery Bay on the north coast of Jamaica, with the bauxite loading terminal situated in thesouth-west corner of the bay. (b) Contour map showingvariations in carbonate content of sediments. Lowest levels occur near the loading terminal and are closely correlated withlevels of iron (c), zinc and manganese. (Adapted from Perry &Taylor 2004.)

ESC09 9/7/06 3:21 PM Page 340


Case Study 9.4 Long-term persistence of oil residues in mangrove sediments, Panama

In April 1986, rupture of an oil store spilled between 75,000 and 100,000 barrels of medium-weight crude oil, at Bahiá las Minas on the Caribbean coast of Panama (Case Fig. 9.4a).Surveys conducted 2 months after the spill indicated that heavy oiling occurred along some 82 km of coastline between Isla Margarita and Islas Naranjos, which harbours around 16 km2

of mangroves and 8 km2 of coral reefs. The only areas along this section of coast to escapeheavy oiling were two restricted mangrove lagoons (Case Fig. 9.4a). The oil had a significantnegative impact upon the mangrove communities, as well as upon nearshore reefs and seagrassinfauna (Jackson et al. 1989). Long-term (> 5 yr) monitoring of mangrove sites along the coastdemonstrated:1 the long-term persistence of residual oils that accumulated within organic-rich mangrove muds;2 the potential for a range of aromatic hydrocarbon residues to be preserved in the sedimentsand for intermittent release into the nearshore environment;3 the impacts on the mangrove communities themselves.

Initial incorporation of oil into the mangrove sediments was rapid. Oil residues were foundat depths of 20 cm within 6 months of the spill, and are likely to have migrated into the sedimentsvia diffusive processes, through crab burrows and down dead or decaying mangrove root casts.The oil residues, however, showed evidence, at most sites, of rapid compositional changesresulting from a combination of weathering processes (including evaporation), dissolution,microbial degradation and photochemical decomposition. Undegraded oil residues were foundonly at a few sites and these are believed to have been preserved within anoxic sediment zones.Despite the apparently rapid compositional changes that occurred, oil residue concentrationsremained high within surface sediments 5 years after the spill, and at many sites had increasedby an order of magnitude at depths of up to 20 cm (Burns et al. 1994).

Panama

0 km 2

IslaMargarita

PuntaGaleta

BahiaLas

Minas PuntaMuerto

IslasNaranjos

R

R Oil refinery

Coastline affected by heavy oiling

Areas free from heavy oiling

A N F M A N F M A N F M A N F M A N F M86 87 88 89 90 91

Open coast

Channels

Drainage streams

Pro

port

ion

of m

arke

r dow

els

oile

d

1

0

1

0

1

0

(a)

(b)

Case Fig. 9.4 (a) Map showing the location of the Bahiá Las Minas oil refinery and the main areas of coast impacted by the April 1986 oil spill. (Adapted from Guzmán et al. 1991.) (b) Plots showing the periodicity of secondary oiling episodes, asindicated by the presence of oil on marker dowels, in different mangrove environments. Arrows denote periods when increasedimpacts were reported on adjacent coral reef communities (data in Guzmán et al. 1994). These appear to immediately follow each of the secondary oiling episodes. (Adapted from Burns et al. 1993.)

ESC09 9/7/06 3:21 PM Page 341

342 CHRIS PERRY

carbonate production. At many sites, reductionsin live coral cover and associated increases inmacroalgal abundance are more realisticallydiscussed in the context of prolonged and varieddisturbances. This is seen particularly clearly onthe reefs of north Jamaica, which have under-gone major changes in coral community struc-ture over the past 25 years (Liddell & Ohlhorst1993). Coral cover in shallow reef environ-ments has declined from around 60% to 2–3%at present, and this has been accompanied by a marked increase in the cover of non-calcifyingmacroalgae. The causes of this decline have beenattributed to hurricane disturbance, coral disease

and coral bleaching, eutrophication, mortalityof the key herbivore Diadema antillarum, and a prolonged history of extensive overfishing.The last two factors, in particular, have resultedin the removal of most herbivorous species, and as a result algal species effectively now out-compete corals. Although the impacts on netrates of carbonate production have not yet beenquantified (but are likely to be significant), shiftsin carbonate sediment production patterns haveoccurred (Perry 1996).

Similar levels of reef decline to those inJamaica have also been recorded along the westcoast of Barbados, and again reflect a varied

Oiling had a significant impact on the mangrove communities over the 5 years to 1991. On the open coast, area of mangrove was reduced by 13% and root density was reduced by24%. In channel and lagoon sites, the area of fringing mangroves reduced by 23% (root densitydecreased by 20%), and in drainage streams, mangrove area reduced by 56.5% (root densitydeclined by 16%) (Levings & Garrity 1994). Release of dissolved and suspended oil residuesfrom mangrove sediments appears to be a relatively frequent occurrence and has led to ongo-ing, pulsed phases of recontamination long after the initial spill. This secondary oiling occursfollowing erosion of mangrove substrates and is most likely related to storm activity, periods of increased rainfall leading to sediment washout, or to the cutting and degradation of areas ofdead mangrove. This pulsed release of oil from the sediments is evidenced by the occurrence of localized oil slicks (which are most common adjacent to those sites most heavily impactedafter the oil spill) and by the re-oiling of artificial marker stakes within the mangroves (CaseFig. 9.4b). This release of oil demonstrates the potential for mangrove sediments to act as long-term storage sites of oil and to periodically release it into the adjacent environments, thus prolonging the time-scales over which oil has an impact on nearshore environments. Time-scales of toxin persistence in mangrove sediments after heavy oiling are estimated to be at least 20 years (Burns et al. 1993).

Relevant reading

Burns, K.A., Garrity, S.D. & Levings, S.C. (1993) How many years until mangrove ecosystems recover fromcatastrophic oil spills? Marine Pollution Bulletin 26, 239–48.

Burns, K.A., Garrity, S.D., Jorissen, D., et al. (1994) The Galeta oil spill: II. Unexpected persistence of oiltrapped in mangrove sediments. Estuarine, Coastal and Shelf Science 38, 349–64.

Guzmán, H.M., Jackson, J.B.C. & Weil, E. (1991) Short-term ecological consequences of a major oil spill onPanamanian subtidal corals. Coral Reefs 10, 1–12.

Guzmán, H.M., Burns, K.A. & Jackson, J.B.C. (1994) Injury, regeneration and growth of Caribbean reef coralsafter a major oil spill in Panama. Marine Ecology Progress Series 105, 231–41.

Jackson, J.B.C., Cubit, J.D., Keller, B.D., et al. (1989) Ecological effects of a major oil spill on Panamaniancoastal marine communities. Science 243, 37–44.

Levings, S.C. & Garrity, S.D. (1994) Effects of oil spills on fringing red mangroves (Rhizophora mangle): lossesof mobile species associated with submerged prop roots. Bulletin of Marine Science 54, 782–94.

ESC09 9/7/06 3:21 PM Page 342


history of recent disturbances, including eutro-phication and storm damage (Lewis 2002). Atthese sites, coral cover is reduced but, in addi-tion, recent work has demonstrated large-scaletemporal change in the areal extent of reefdevelopment. Three main types of structuralloss are identified:1 degradation of spur-and ‘groove’ structureson the seaward edge of reefs;2 breaches in the reef front, leading to thedevelopment of sand-filled valleys;3 damage and loss of flank areas.These sites provide clear evidence for reducedreef development (Fig. 9.21) and changes in carbonate production that can be attributed torecent natural and anthropogenic disturbanceevents. These types of change, where coral-dominated reefs have shifted to algal-dominatedcommunities, are far from isolated examples,and conform to the ‘phase shift’ concept of Done(1992). Whether these changes are permanent

or whether they represent temporary shifts incommunity state remains unclear.

9.5 MANAGEMENT AND REMEDIATION OF CORAL REEF

AND MANGROVE SEDIMENTS

Coral reef and mangrove environments are susceptible to a range of disturbance events thatnecessitate increasing attention from a manage-ment and remediation perspective. In part, thisrelates to the effects of anthropogenic-induceddisturbance and related ecological degradation,which can in turn lead to destabilization of sedi-ment substrates and increased rates of (shoreline)erosion. Most coastal systems, however, are alsosubject to shoreline change induced by naturalfluctuations in energy inputs (such as thoselinked to storms, cyclones or tsunami). In thesecases, the need for effective management is oftendriven not so much by the actual event, but as a

Bridgetown

Holetown

Speightstown

BARBADOS

Reefs

0 (km) 10

N

Coastline

0 m 100

0 m 100

(b) Paynes Bay

(a) Six Mens Bay

Extent of reef development

19501991

Coastline

Fig. 9.21 Structural changes on fringing coral reefs on the west coast of Barbados between 1951 and 1991. At Six Mens Bay (a)recession of most of the reef perimeter has occurred, although isolated outliers of live coral remain. This has resulted in a 24% reductionin reef area. At Paynes Bay (b) major loss on the northern flank and in the central region has occurred, leading to a 21% reduction in reefarea. At both sites the areas of reef loss now comprise sand and rubble. (Adapted from Lewis 2002.)

ESC09 9/7/06 3:21 PM Page 343

344 CHRIS PERRY

result of (increasing) human occupation and useof the coastal zone, i.e. urbanization of environ-ments that will naturally respond to changes innearshore energy levels. Extensive discussion ofthe ecological aspects of, and problems relatedto, reef and mangrove management are beyondthe scope of this chapter, but useful reviews areprovided by Field (1998), Gladstone et al. (1999)and Yap (2000). Rather, this section outlines anumber of key management issues relating toshoreline change, coastal sediment budgets andsediment contamination.

9.5.1 Managing natural coastal hazards

Given the naturally dynamic nature of manymangrove and reef-fringed shorelines (see sec-tion 9.3.1), understanding the effects of flood-ing and wave-induced erosion represents animportant issue in tropical coastal management.These impacts can be caused both by high-magnitude events, such as cyclones and tsunami,as well as ongoing or periodic processes linkedto land subsidence or El Niño-type events. In all cases, resultant changes in nearshore energy levels alter the natural equilibrium of the systemand generate a geomorphological response. Thismost commonly occurs in the form of eithershoreline erosion or sediment remobilizationand deposition. In essence these sedimentaryenvironments respond to the new (albeit, inmany cases, temporary) physical conditions bychanging their equilibrium state.

The potential impacts and management im-plications of such geomorphological changeshave been discussed for a range of differentSouth Pacific coastlines (high volcanic islandsand atolls) by Solomon & Forbes (1999). In allcases, significant shoreline erosion and floodingoccurred, although marked spatial variationsoccur in the extent of erosion. In some casesthese variations related to natural differences in nearshore geomorphology and the extent ofreef development (which influence the loca-tion and height of breaking waves), whereas inothers erosion and flooding were exacerbatedby human modification of the nearshore sys-tem. These include the removal of nearshoresands during mining, the use of inappropriate

shoreline protection schemes, and the destruc-tion of coastal vegetation. Reclaimed land appearsparticularly susceptible to erosion. In all cases,economic damage was linked to inappropriatelocation of roads and buildings in areas that can be expected to be impacted by these types of disturbance.

Natural hazard assessment and considerationof nearshore sediment dynamics therefore forman important aspect of tropical coastal manage-ment and an important component of IntegratedCoastal Management schemes. Historical recordsof physical disturbance and shoreline changecan provide an insight into the location of high-risk areas and of sites subject to either severe or episodic sediment erosion. Such risks maynecessitate the delineation of coastal zones thatare deemed inappropriate for subsequent develop-ment, although this will present an increasingproblem in areas with restricted coastal plainsand rapidly increasing populations (Aubanel et al. 1999).

9.5.2 Managing resource extraction andexploitation

Common factors contributing to the magni-tude of damage associated with natural dis-turbance events and, in other cases, direct causesof shoreline erosion (see section 9.4.1), are highlevels of sediment extraction (due to mining and dredging), degradation of reef structures(due to coral mining and channel blasting) andremoval or loss of intertidal areas. In most cases,subsequent shoreline change (erosion, channelsiltation) and a magnification of the effects ofepisodic high-energy events can be linked tochanges in nearshore sediment budgets, cur-rent dynamics and artificial modification of thecoastal fringe. In addition to limiting or prevent-ing resource extraction, rehabilitating degradedenvironments is often costly and problematic.The rehabilitation of directly degraded man-grove swamp previously claimed for farmingwill, for example, necessitate reinstating con-ditions in relation to topography (and thus tidal inundation), hydrology, sedimentationrates and sediment characteristics. The formermay be possible via land regrading and artificial

ESC09 9/7/06 3:21 PM Page 344


modification of channel networks (Lee et al.1996). Restoring the physical and chemical prop-erties of mangrove sediments, however, mayprove impossible at sites that have been used for shrimp farming or timber harvesting, whereextreme acidification and anaerobiosis may haveoccurred. If devoid of vegetation, these areas mayalso be prone to rapid soil erosion and removalof organic matter (Field 1998). Under such con-ditions, natural reseeding or programmes ofreplanting are unlikely to be successful.

The need for effective regeneration of degradedcoral reefs is equally evident both from the perspective of disturbance-related reductions incarbonate production and from the increasedrates of adjacent shoreline erosion that can follow reef disturbance (see section 9.4.1).Remedial techniques such as coral transplanta-tion do exist, but as with mangrove replanting,the benefits of this approach are questionable.Edwards & Clark (1998) suggest that naturalrecovery processes may be far more effectivethan transplantation, with the exception of areasthat are failing to recruit juvenile corals. Even inthese cases they suggest a re-emphasis towardstransplanting slow-growing, massive corals,which tend to be better natural recruiters thanfast-growing branched corals. Massive coralsare also less susceptible to the effects of coralbleaching (McClanahan 2000). At best, trans-plantation is only likely to be effective on alocalized scale, so that resources may be betterdirected at managing or mitigating the causes ofdegradation (Bellwood et al. 2004).

9.5.3 Managing sediment contaminants

As outlined in section 9.4, nearshore sedimentsin general, and mangrove sediments in par-ticular, have considerable potential to trap a wide range of contaminants relating to indus-trial, urban and agricultural discharges. In mangroves, these contaminants, if present in sufficient concentrations, can lead to tree defoliation, seedling mutations, reduced speciesdiversity and mangrove die-back (Ellison &Farnsworth 1996), and in coral reefs to disease,and reduced growth and reproductive potential.In many coastal systems a range of contamin-

ant inputs can often be identified, some beingsourced from the coastal fringe, but manylinked to activities that occur upstream withincoastal catchments. The diverse range of bodieswith responsibilities to regulate these different,but linked environments can cause significantmanagement problems.

The central Great Barrier Reef (GBR) providesan interesting perspective on the uncertaintiesthat exist in relation to terrigenous impacts onnearshore marine environments, and the issuesof coastal and marine management. There have,for example, been suggestions of increasing sedi-ment runoff and, associated with this, increasednutrient and contaminant (pesticide) levels,linked to agricultural activities in the catchmentareas (Haynes & Johnson 2000). Although sedi-ment input rates do appear to have increased,there is conflicting evidence about whether ornot this is having a detrimental effect on themarine environment and on the coral reefs inparticular, because most fluvially derived sedi-ments accumulate only along nearshore areas ofthe shelf (Larcombe & Carter 2004 – see CaseStudy 9.1). The GBR region, however, does high-light the important linkages that exist betweencoastal environments and river catchments.Management of the GBR is primarily under thejurisdiction of the Great Barrier Reef MarinePark Authority (GBRMPA). Although the GBRitself is widely cited as an example of goodmarine management practice, the GBRMPA haslittle influence over managing the catchmentsthat feed onto the shelf. Instead, current man-agement of agricultural activities and hencerunoff are under a voluntary code of practice.Given the uncertainties about the actual impactsof terrestrial sediment inputs on the GBR, theeffect of changing land-use practices, in terms of the marine environment, seem unclear. Theregion, however, emphasizes the potential bene-fits of integrating both coastal and catchmentmanagement schemes.

Remediation of contaminated sediments pre-sents an additional challenge and one that isimportant given the potential for sediments toboth store and episodically re-release contamin-ants (see section 9.4.4). In the case of oil con-tamination, chemical dispersants traditionally

ESC09 9/7/06 3:21 PM Page 345

346 CHRIS PERRY

have been used to try and mitigate the effects of oiling. Evidence suggests that although dis-persants may be useful in limiting the amount of oil that reaches mangroves, and may limitmangrove tree mortality, they are less effectiveat preventing oil absorption into mangrove sedi-ments, or subsequent patterns of oil weatheringand degradation (Duke et al. 2000). Some suc-cess in remediation of oiled sediments has beenachieved through the use of bioremediationtechniques (Ke et al. 2003). These exploit micro-organisms that occur naturally within mangrovesediments, some of which are effective degradersof hydrocarbons. Although rates of bacterial oil degradation may be inhibited under normalanaerobic sediment conditions, rapid degrada-tion rates have been achieved by manipulatingbacterial population densities via active sedimentaeration and the application of slow-release fertilizers (Ramsay et al. 2000).

9.6 EFFECTS OF GLOBAL CLIMATIC AND ENVIRONMENT

CHANGE ON TROPICAL COASTAL SYSTEMS

Many of the predicted changes in global climaticconditions (see Chapter 1) have the potential toinfluence both coral reef and mangrove environ-ments. Those with direct implications includechanges in atmospheric CO2 concentrations,increased atmospheric and sea-surface tempera-tures, increased UV radiation, changes to patternsof storm frequency and intensity, and increasedsea-level (Ellison & Farnsworth 1996; Wilkinson1996). Depending upon the magnitude of thesechanges and the local sedimentary regime, coralreefs and mangroves may exhibit highly vari-able, and often site-specific, responses. A crucialpoint to emphasize is that environmental changemay have both positive and negative effects uponcoral reefs and mangroves.

9.6.1 Impacts of climate change on coral reef andmangrove sedimentary systems

Changes in atmospheric CO2 levels may haveboth direct and indirect effects on coral reefsand mangroves. In mangroves, increased CO2

levels are likely to enhance rates of photo-synthesis leading to increased leaf production and mangrove growth (Ellison & Farnsworth1996), although the positive effects of increasedproductivity may be restricted locally by sea-level rise (see section 9.6.2). Elevated CO2 levelsmay have an impact on coral communities bymodifying marine aragonite saturation statesand thus reducing rates of coral calcification(Kleypas & Langdon 2002). Recent modellingstudies through to 2069 predict that much of thePacific basin may become marginal (in terms of calcification potential) for corals with respectto aragonite saturation (Guinotte et al. 2003)and would have major implications for reef carbonate production.

Linked to increasing CO2 levels are probableincreases in atmospheric and sea-surface tem-peratures. In most mangrove settings predictedtemperature increases are not likely to be highenough to cause direct mangrove mortality(Ellison & Farnsworth 1996) and, in higher lati-tude regions, may actually promote expansionof mangrove swamps. Such expansions wouldbe dependent upon local hydrological constraintsand, in arid settings, conditions may actuallybecome increasingly marginal for mangrovesurvival. Higher atmospheric temperatures andincreased solar radiation may also cause sedimentwarming, leading to increased soil respiration,organic matter decomposition, methane (CH4)and hydrogen sulphide (H2S) release, and rootturnover. This would, in turn, modify the com-position and chemistry of mangrove sediments,although the precise impact on mangrove com-munities is not clear. In contrast, coral reef communities are likely to be influenced signi-ficantly by increasing sea-surface temperatures(Chadwick-Furman 1996). From a positive per-spective such warming may permit an increase in the latitudinal range of hermatypic corals, and phases of reef ‘switch-on’ in higher latitudecarbonate shelf settings, whereas a negative con-sequence may be increased coral bleaching. Thisoccurs as a physiological response to stress, andinvolves coral shedding of the symbiotic algae,leading to reduced coral growth and, if conditionspersist, mortality (Glynn 1996). The effects of

ESC09 9/7/06 3:21 PM Page 346


such increases were evident most recently on aregional scale following the strong El Niño of1997–98, which resulted in widespread coralmortality across the Indo-Pacific. Such bleaching-related mortality has major consequences forreef carbonate production, as has been quanti-fied in Panama (Eakin 1996; see section 9.3.3).Temperature-related bleaching events have alsobeen documented in higher latitude settings,where corals are adapted to lower average temperatures (Celliers & Schleyer 2002) andhence the effects of such bleaching events arelikely to be globally significant. The rate atwhich coral communities will recover fromthese widespread bleaching episodes remainsunclear, but may depend upon the ability ofcorals to adapt to increasing mean sea-surfacetemperatures.

Shifts towards net erosional regimes on manyshallow reefs are thus a potential consequence ofincreased bleaching. This may lead to destabil-ization of reef frameworks and reduced structuralintegrity, although the response of frameworkaccumulation processes (such as bioerosion

and encrustation) to such disturbances remainspoorly documented. Any suppression of car-bonate production will, however, inhibit thepotential for reefs to respond positively to sea-level increases (see section 9.6.2) and increasethe frequency of storm-wave overtopping. This,in turn, may result in changes in nearshore sediment dynamics and changes in shorelinemorphology (Fig. 9.22). Climate change mayalso modify storm patterns and intensities, andshift the position (or latitudinal range) of thecyclone/hurricane belt, and thus influence reefand mangrove structure (Smith et al. 1994). Inaddition to any increase in physical disturbance,more frequent/intense storms may also have animpact on mangroves by increasing rainfall andsediment runoff.

9.6.2 Impacts of sea-level change on coral reef andmangrove sedimentary systems

A major consequence of global climatic changewith implications for coral reef and mangroveenvironments is sea-level change. Future trends

Increased atmospheric CO2 levels

Increased air temperatures

Increasedstorm frequency

and intensity

Increasedsea-surface

temperatures

Increased sea-level

Coral bleaching

Renewed reef growth

Increased coraland mangrove damage(+ structural changes)

Reduced CO32− concentrationsand reduced

coral calcification

Degradation and destabilizationof reef structures and changing rates

and patterns of carbonate production?

Increased photosythesis and leaf production

promoting mangrove growth

Sediment warming leading to increased

organic matter degradation, CH4 + H2S release and root turnover

Changes in sediment chemistry? Mangrove expansion of low-lying areas?

orSubstrate destabilization and mangrove mortality?

Increased tidalinundation

New coral growth in high latitude areas?

Promote mangrove growth/expansion?

orMangrove mortality and

substrate destabilization?

Fig. 9.22 Simplified schematic diagram illustrating some of the key processes associated with climate change (boxed) and bothpositive (bold italics) and negative (italics) responses within coral reef and mangrove environments. Note that many of the potentialresponses have questions marks, highlighting either the uncertainties that exist, or the likely site-specific nature of the responses.

ESC09 9/7/06 3:21 PM Page 347

348 CHRIS PERRY

(and uncertainties) regarding sea-level are outlinedin Chapter 1, but mean sea-level rises of around50 cm by 2100 are predicted (Wilkinson 1996)and equate to annual increases of around 4 mmyr−1. In the context of published coral growthrate data, which can be as high as 10–12 mmyr−1, these predicted changes thus appear rela-tively insignificant. Actual reef accretion rates(the rate at which the reef grows vertically) aredetermined not only by coral growth, however,but also by factors such as the rate of bioerosionand sediment production, and actual rates of net reef accretion are far lower (Stoddart 1990).In addition, reef accretion rates vary spatiallyacross individual reef systems and so differentreef subenvironments (reef flat, shallow reef-front, deep reef-front) will have varying potentialto maintain their position relative to sea-level(Spencer 1995).

Depending upon the time-scales over whichreef accretion trends are viewed, it is possible to reach very different conclusions about theresponse of reefs to sea-level change. Recentaccretion rates suggest that although those areasof reef that are characterized by fast-growingbranched corals can accrete at rates that exceedall but the highest sea-level rise estimates (Fig. 9.23a), accretion rates for reefs as a whole(i.e. encompassing lagoon through to deep reef-front environments) fall below even the lowersea-level rise scenarios (Spencer 1995). In con-trast, longer term (geological time-scale) datafrom coring studies indicate that many reef systems have accreted at rates close to or abovethose predicted over the next century (Fig. 9.23b).Such data, however, include growth that occurredduring the early Holocene when sea-levels wererising rapidly (around 5 mm yr−1 between 10.5and 7.7 kyr BP in the Caribbean; Toscano &Macintyre 2003) and it is unclear how modernreefs would respond to such changes. Based onHolocene reef accretion records the thresholdrates beyond which reefs are unable to main-tain pace with sea-level change lie at around 8–10 mm yr−1 (Spencer 1995). One benefit ofsea-level rise may be stimulation of reef growthin areas that have reached sea-level and stabilizedover the past 5000 years (Wilkinson 1996).

Linked to the response of coral reefs to sea-level rise is the question of how low-lying car-bonate islands and rubble cays may respond.These islands comprise semi-consolidated sandand rubble sitting atop reef platforms, and com-monly perceived threats include direct erosion(or submergence) and increasing saltwater intru-sion into island aquifers (Wilkinson 1996). Recentmodelling studies, however, suggest consider-able complexity in island response to sea-levelrise, related in large part to changes in sedimentsupply (Kench & Cowell 2002). Reduced sedi-ment supply, combined with increased sea-levels,may result in increased shoreline displacementrates on atolls. Reef flats typically produce littlesediment, however, and the most significant dis-placement rates are caused by changes in the littoral sediment budget. These may occur as the result of a range of additional anthropogenicstressors (see section 9.6.3) and thus magnifythe impacts of sea-level change.

The ability of mangrove shorelines to maintaintheir current positions and geometries in responseto sea-level rise appears equally dependent uponthe rate of sea-level change and local sedimentdynamics, and is likely to be highly site-specific.Stratigraphical studies through Holocene man-groves suggest that many of the large modernmangrove swamps did not exist during the earlyHolocene (Woodroffe 1990). This is attributedto the rapid rates of sea-level rise (in the order of 10 mm yr−1) and to differences in shorelinegeomorphology that prevented sediment accu-mulation and restricted mangroves to isolatedfringing communities. Research indicates thatmodern mangroves are likely to be eroded if sea-level rises exceed 0.8–0.9 mm yr−1 (Ellison &Stoddart 1991), but this will depend upon localsedimentation rates.

Mangroves that currently develop on lowlying carbonate islands, which are character-ized by autochthonous sedimentation and typic-ally exhibit low accretion rates (< 0.8 mm yr−1),may experience either expansion or contractiondepending on local rates of sediment productionand accumulation (Ellison & Stoddart 1991). Incontrast, both river- and tide-dominated man-groves can potentially receive large amounts of

ESC09 9/7/06 3:21 PM Page 348


(a)

(b)

High

Middle

Low

Maximum

High (coral thickets)

Intermediate (Reef flat)

'Complete' reef system

10

9

8

7

6

5

4

3

2

1

0

Rat

e of

sea

-leve

l ris

e(m

m y

r−1)

and

rate

s of

ree

f acc

retio

n (m

m y

r−1 )

High

Middle

Low

14

13

12

11

10

9

8

7

6

5

4

3

2

1

0

Rat

e of

sea

-leve

l ris

e(m

m y

r−1 )

and

ra

tes

of r

eef a

ccre

tion

(mm

yr−

1 ) Frin

ging

/Bar

rier

reef

s

Indo

-Pac

ific

atol

ls

Nor

ther

n G

BR

Cen

tral

GB

R

Sou

ther

n G

BR

Bel

ize

Bar

rier

Ree

f

1990–2030 1990–2070 1990–2100

1990–2030 1990–2070 1990–2100

0.81.1

1.4

3.34.0

4.4

6.3

7.58.2

0.81.1

1.4

3.34.0

4.4

6.3

7.58.2

Fig. 9.23 (a) Rates of modern reef accretion for different reef subenvironments and for reef systems as a whole compared with low,medium and high projections of global sea-level rise for three different time periods through to 2100. The different net accretion ratesthat occur in different reef subenvironments reflect differences in rates of coral growth and coral substrate degradation (see Perry 1999).(b) Rates of Holocene reef growth based on radiocarbon-dated drilled core-sections superimposed above low, medium and highprojections of global sea-level rise for three different time periods through to 2100. The often high average rates of accretion during theHolocene reflect growth under different regimes of sea-level rise during the Holocene transgression. Projections based on IS92aemissions scenario. (Adapted from Spencer 1995.)

ESC09 9/7/06 3:21 PM Page 349

350 CHRIS PERRY

sediment and so might be expected to respondmore positively. These environments, however,are highly variable in form and are characterizedby very different sediment dynamics and accre-tion rates. River-dominated mangroves with highsediment inputs may accrete at rates that are suf-ficient to keep pace with sea-level rise (Woodroffe1990). Those associated with smaller river catch-ments (with lower sediment influx) or areas sub-ject to seasonal sediment influx may undergolandward migration. Tide-dominated mangrovesettings also vary significantly in terms of tidalinfluence and sediment flux and, in such settings,shoreline response will again be highly site specific(Wolanski & Chappell 1996).

Predicting mangrove response to sea-level riseis further complicated by the fact that futuresea-level rises will occur not only from condi-tions of relative stability (Woodroffe 1990), butalso across relatively low-lying land (where thespatial extent of sea-level increases is signific-antly higher). Although this will increase therate at which tidal inundation and associatedchanges in sediment characteristics and salinitylevels occur (Semeniuk 1994), from a positiveperspective it may also facilitate landward mig-ration of mangroves. Indeed, in cases where sediment accumulation rates are high and theseaward margins of the swamps are maintained,the spatial extent of mangrove swamps mayactually increase. This, however, will be inhib-ited in areas where urbanization of the coastalfringe has occurred and under these conditions,regardless of sediment influx, progressive man-grove degradation is likely.

9.6.3 Climatic and sea-level change in relation toincreased anthropogenic influence

Although global climate changes and relatedincreases in sea-level are often discussed fromthe context of increasing ecosystem disturbance,there is an important caveat to such speculation.Predicted changes in temperature, CO2 con-centrations and sea-level through to 2100 areunlikely to approach the scales of change thathave occurred in the recent geological past

(Wilkinson 1996). For example, major changesin sea-level (of 100+ m) occurred at varioustimes during the Pleistocene. In addition, thereis some evidence to suggest that Holocene sea-levels underwent rapid ‘jumps’ at several pointsover the past 16,000 years (Toscano & Macintyre2003) and that rapid rates of sea-level rise (up to20 mm yr−1) characterized the early Holocene(Wilkinson 1996). Neither appear to have hadlong-term detrimental effects on reef or man-grove communities, although massive changesin geomorphology and structure have obviouslyoccurred as the world’s sedimentary systems mig-rated across the continental shelves. Similarly,atmospheric CO2 concentrations and temper-atures have fluctuated significantly through the Pleistocene (Wilkinson 1996) without anyapparent long-term negative impact on reef andmangrove communities.

Over the coming century, different sedimen-tary responses may, however, result from thecombination of natural variation and anthro-pogenic influence. Although coral and mangrovecommunities and their associated sedimentaryenvironments are capable of responding to natural changes in physical and environmentalfactors, future responses may be influenced bythe additional stresses imposed by a range ofanthropogenic-related factors. In coral reefs,modified rates and patterns of reef carbonateproduction may, for example, occur as a resultof changes in reef community composition drivenby a range of factors including overfishing andpollution. Similarly, in mangrove swamps, mangrove mortality and substrate erosion mayoccur due to sediment contamination and land-use change. These additional pressures mayalter the ways in which reef and mangrove sys-tems respond to climatic and sea-level change.Our understanding of how both reef and man-grove systems are influenced by specific anthro-pogenic stressors, the mechanisms and rates of recovery and adaptation potential, and thelonger term impacts on rates of carbonate production and accretion (in coral reefs) andsediment accumulation (in mangroves) remainpoorly understood.

ESC09 9/7/06 3:21 PM Page 350

10.1 INTRODUCTION

This chapter addresses the environmental sedi-mentology of continental shelves. ‘Continentalshelf’ is taken to be the sea floor shallower than about 200 m adjacent to continental land-masses, with the outer margin marked by thecontinental slope (e.g. Whitten & Brooks 1972)(Fig. 10.1). In the long-term, continental shelvesreceive river sediments, with supply to the shelf modulated by catchment (Milliman 2001;Walling & Fang 2003) and estuarine processes(Dyer 1966, 2000). Some shelves contain majorregions of in situ sediment production, pre-

10Continental shelf environments

Piers Larcombe

dominantly calcareous in nature, and shelf sediment dynamics are crucial in influencing the nature and fate of shelf sediments. Thenature and morphology of continental shelvesare controlled by: (i) the hydraulic regime, andhence sediment transport, (ii) sediment supplyand (iii) relative sea-level (Johnson & Baldwin1996). Other factors, particularly important forcarbonate shelves, are climate, biological inter-actions with the sediments, seawater chemistryand sediment composition. It is increasinglybeing acknowledged in management regimesthat the surface sedimentary systems are vital tomarine ecosystems.

129

1110

5

72

4

1

3

6

8

13

Fig. 10.1 Distribution of modern continental shelves (dark grey). Continental shelves noted in the text are: 1, Great Barrier Reef,Australia; 2, Western New Caledonia; 3, Lacepede shelf, South Australia; 4, South Otago, New Zealand; 5, Amazon; 6, Belize; 7, south-east Africa; 8, southern Namibia; 9, Oregon–Washington; 10, Texas–Louisiana shelf; 11, Mid-Atlantic Bight; 12, UK and westernEuropean shelf; 13, north-western Mediterranean. (Modified after Johnson & Baldwin, 1996.)

ESC10 9/5/06 3:09 PM Page 351

352 PIERS LARCOMBE

Direct human influences on shelf sedimentarysystems probably can be constrained mainly to the past few hundred years, but such a viewrequires assessing the nature of environmentalchange over the longer time-scales of many ofthe natural driving processes. Further, the sedi-mentary environmental setting is vital. Thischapter thus describes the geoscientific contextand sedimentary processes of continental shelvesof at least the past few hundred years, noting the influence of human use and management. Adeliberate focus on the Australian and north-west European continental shelves introducesthe reader to two contrasting but well-describedregions. Compared with Chapters 7, 8 & 9, thefocus herein is those areas of the shelf where thedominant natural inputs of energy and materialare not land-influenced, and where environmentalmanagement requires understanding marine processes. There is no attempt to be exhaustive,continental shelf settings are also discussed byWalker & James (1992), Wright (1995), Johnson& Baldwin (1996), Wright & Burchette (1996)and Allen (1997).

10.1.1 Nature of shelf sedimentary environments

The Holocene sea-level highstand and the asso-ciated relatively deep, modern continental shelvesare the geological exception rather than the rule(Shackleton et al. 1990). As a result, modernshelves have significant ‘accommodation space’(the total space below sea-level within which itis possible to deposit marine sediment), primarycontrols on which include relative sea-level andshelf topography. Modern continental shelvesare subject to a wide range of physical and bio-logical sedimentary processes, but these are ofvaried relative significance. Considering phys-ical processes, Swift et al. (1986) concluded that80% of modern continental shelves are stormdominated, 17% are tidally dominated and only3% are dominated by oceanic currents. On high-energy, exposed shelves, long-period waves canmobilize sediment even in fair-weather condi-tions, whereas a complex suite of shelf currents,including waves, wind-driven currents and ‘return’currents, can occur during storms (Wright 1995).Thus, a grouping can be made based on the

Storm-dominated

Wave-dominatedTidal-dominated

Oceanic current-dominated

Fair-weather processes1

2

3

4

5

Tides increasingly dominant fair-weather process

Oscillatory flows increasingly dominant fair-weather process

Tida

l pro

cess

es in

crea

sing

ly

dom

inan

tSt

orm

pro

cess

es in

crea

sing

ly

dom

inan

t

Increasing dominance of

fair-weather oscillatory flow

s

Increasing frequency of

intense oscillatory flows

Incr

easi

ng th

erm

o-ha

line

curr

ents

Incr

easi

ng w

ind-

driv

en c

urre

nts

Stor

m c

hara

cter

istic

sH

igh

inte

nsity

and

mod

erat

e–

high

inte

nsity

(hur

rican

e/

tropi

cal s

torm

bel

t?)

Mod

erat

e–hi

gh in

tens

ity

and

low

–mod

erat

e fre

quen

cy

(win

ter s

torm

bel

t?)

Low

–mod

erat

e

inte

nsity

and

low

frequ

ency

Fig. 10.2 Classification of modern continental shelves based on physical processes: 1, tidal straits; 2, tidal seas; 3, oceanic-currentswept shelves; 4, storm-dominated shelves; 5, river mud-dominated shelves. (Simplified after Johnson & Baldwin 1996.)

ESC10 9/5/06 3:09 PM Page 352

CONTINENTAL SHELF ENVIRONMENTS 353

dominant fair-weather processes and the rela-tive interaction with storm-associated processes(Fig. 10.2), where of primary interest are thoseprocesses that are reflected in the nature, dis-tribution and stratigraphy of shelf deposits.Shelf sedimentary processes (Fig. 10.3) operateat a range of physical and temporal scales, sothat the relative sedimentary impacts of theseprocesses depend on their frequency, magnitudeand recurrence intervals. Sediment transport is a cubic (or quartic) function of flow speed(Soulsby 1983, 1997), which, together with the

presence of flow thresholds of transport, meansthat short periods of fast flows (e.g. related tostorms, cyclones or tsunami) may sometimes bemore significant in terms of sediment transportthan longer periods of slower flow.

10.1.2 Environmental sedimentology of shelfenvironments

Human influences on the sedimentary regime ofcontinental shelves include, but are not limitedto, impacts by fisheries (especially trawling),

0.1 mm 1 mm 1 cm 10 cm 1 m 10 m 1 km 10 km 100 km 1000 km

Spatial scale

Tem

pora

l sca

le

105 yr

10,000 yr

1000 yr

100 yr

10 yr

1 yr

1 mo

1 wk

1 day

10,000 s

1000 s

100 s

10 s

1 s

0.1 s

0.01 s

Suspensionbands

Ripples

Dunes

Beds

Depositionalsystems

Depositionalsystems

tract

Laminae

Nearshoremorphology

Facies

Events: storm surges, flooding,water level fluctuations due to seiches, river run-off

Forcing functions

Sediment characteristics

Millenia: macro-scalesediment balances, sea-level fluctuations, climate, neotectonics

Decades: storm beddeposition

Centuries: macro-scale sedimentbalances, wind climate, sea-level fluctuations

Seasons: shoreface variability,calm/stormy weather,river run-off

Fig. 10.3 Forces operating in coastal and shelf environments and their sedimentary products. (Modified from Sternberg & Newell1999; Schwarzer et al. 2003.)

ESC10 9/5/06 3:09 PM Page 353

354 PIERS LARCOMBE

mineral extraction, disposal of dredged mater-ial and changed supply of material from rivers.Sediment supply might be either reduced be-cause of damming of large rivers or increased asa result of deforestation in catchments leadingto higher sediment yields, or by the canalizationof some rivers, which leads to increased bypass-ing of sediment through the coastal zone. Formany types of human impact on the sediment-ary environment, the main method of measur-ing change is restricted to historical surveys, but consideration of the significance of humanimpacts means that it is necessary to measurechange on time-scales of decades to centuries(Berger & Iams 1996; Sarnthein et al. 2002). Ofparticular interest are those radiometric tech-niques that allow delineation of chronologiesover a few decades (e.g. 210Pb, 137Cs and others).On continental shelves themselves, the best re-cords of past change are generally formed atsites with high rates of sediment accumulation(e.g. the Amazon shelf, Kuehl et al. 1986) or on shelves that have biogenic features with ahigh preservation potential, from which can bederived a proxy record of environmental change(e.g. Schöne et al. 2004).

Sediments form habitats for marine fauna andflora, and one view of environmental manage-ment is that the overall mangement aim shouldbe habitat retention, on the basis that if thehabitats themselves are present, then it is likelythat the structure and function of the associatedecosystem components will also be retained. Shelfsediments may accumulate contaminants to theextent that the biology is negatively impacted,and these surface accumulations may them-selves become sources of pollution, with regardto transport across the shelf and/or bioturbationdown into the sediment column (e.g. Kershaw et al. 1988; MacKenzie et al. 1999). Trace metalelements are often of environmental concern,background concentrations of which are con-trolled by the geological history of a continentalshelf and its catchments (e.g. Chapter 1). Thereare few texts regarding human management ofshelf environments, and much informationderives from reports of past or ongoing activ-ities. The websites of Government departments

and associated regulatory bodies are increas-ingly useful. Examples include the US Environ-mental Protection Agency (EPA), EnvironmentCanada, the UK Department of Environment,Food and Rural Affairs (DEFRA), the Nether-lands National Institute for Coastal and MarineManagement, the Australian Department ofEnvironment and Heritage, the Great BarrierReef Marine Park Authority (GBRMPA), theNew Zealand Department of Conservation andNational Institute for Water and Atmosphere.

10.2 SHELF SETTINGS AND SEDIMENT SOURCES

10.2.1 Shelf settings

Continental shelves receive sediments that havebeen transported from the terrestrial realm byrivers, are the sites of biogenic sediment produc-tion, store sediments in various environmentsand for various time-scales, and transport somesediments off-shelf down the continental slopetowards the abyssal plains. There are two mainmorphological types of continental shelf (Johnson& Baldwin 1996):1 pericontinental shelves – which occur on con-tinental margins and equate to those moderncontinental shelves with the classic division and profile of shoreline, shelf and slope (e.g. theAmazon and Californian shelves);2 epicontinental shelves – which are partiallyenclosed seas within continental areas usuallywith a uniform shallow-dipping ramp profile(e.g. the Gulf of Mexico and the North Sea).

Tectonically, the largest continental shelves arelocated at passive margins, where they generallydevelop seaward-thickening accumulations ofsediment, supplied by large continental drain-age systems. At active plate margins, shelves maydevelop at the convergent margin itself, uponwhich shelf areas are relatively small and whichmay have zones with high sediment accumula-tion rates, or in foreland basins, where extensiveareas of continental shelf may develop and sedi-ment accumulation rates may be high.

Understanding the processes and features ofactive shelf sedimentary systems involves long

ESC10 9/5/06 3:09 PM Page 354


time-scales and major changes in sea-level. At thelast glacial maximum (LGM), 20,000–18,000years ago, sea-levels around the globe were120–130 m below present (Chappell et al. 1996)because large volumes of ocean water were lockedup as ice in polar regions. Those physical areasthat now form continental shelves were thussubject to subaerial processes and the lowstandcoastline was located to seaward of the presentcontinental shelves. Inundation of these areasoccurred during the post-glacial transgression,which together with processes operating at themodern sea-level highstand reworked the sub-aerial deposits to various extents. Viewed simply,the post-glacial sea-level rise was rapid untilaround 8000 years ago, when it slowed to reachits highest level around 6000 years ago. Con-sequently, many regions of the world’s contin-ental shelves have been subject to their presentprocesses for only a few thousand years, and manyof the bathymetric, mineralogical and granulo-metric features of continental shelves and theirsediments derive from the actions of past pro-cesses that operated at different stages of relativesea-level. Considerable regional variability occursin relative sea-level, because of the interactionsof ‘global’ sea-level, gravitational variations, tec-tonics, hydro-isostacy and weather, operating atdifferent temporal and spatial scales.

10.2.2 Sediment sources and characteristics

10.2.2.1 Fluvial/lithogenic material

Chemical and physical weathering of rocks in river catchments leads to sediments being carried by rivers to the sea. Globally, the totalload of silt delivered by rivers to the ocean hasbeen estimated to be 13.5 × 109 t yr−1 (Milliman& Meade 1983), to which bedload transportadds 1–2 × 109 t yr−1. The combined rate of sedi-ment supply represents an average denudationof the world’s catchments of 50 mm ky−1 (Allen1997), but there is great geographical variationin rates of erosion and transport, controlled byfactors including the rate of uplift, climate, rocktype, topography and precipitation. Sedimentdelivery also can be affected by human activities

in catchments and rivers, such as deforestationand conversion to agriculture (e.g. Chapters 2& 3). There are therefore wide variations in the nature and volume of sediments delivered to the shelves (Chapters 1 & 8). Some shelvesreceive huge volumes of sediments, particularlyfrom those rivers draining steep tropical catch-ments with relatively young geology, such as theIndonesian Archipelago (Fig. 10.4). The AmazonRiver produces an average of 1.2 × 109 t yr−1 ofsediment, about 90% of which is silt and clay(Dunne et al. 1988), and which represents anaverage denudation rate of the catchment of 69 mm kyr−1. On relatively dry and old contin-ents, such as Australia, the largest river in termsof sediment delivery is the Burdekin. This delivers3 × 106 t yr−1 of sediment to the continentalshelf, the sediment comprising around 75% siltand clay. Sediment yields equate to a catchmentdenudation rate of 9 mm kyr−1 (Belperio 1979;Allen 1997). For some continental shelves thatreceive relatively little river input, materialreworked from the shelf seabed by storms orstrong currents may form a significant com-ponent of the overall shelf sediment budget.

10.2.2.2 Shelf in situ production

Continental shelves produce significant amountsof ‘new’ sediment, mainly biogenically, but also locally by inorganic precipitation in some tropical environments. On some shelves and incertain environmental conditions, authigenicminerals are formed, such as the clay mineralglauconite (Bornhold & Giresse 1985; Odin1988). By far the most significant global con-tribution is that of biogenic material, whichincludes skeletal parts of animals and marineplants, of various calcareous and siliceous com-position, living and growing within the watercolumn and on the sea bed (Walker & James1992; Wright & Burchette 1996). Significantproduction of biogenic sediments needs a suit-able regime of salinity, temperature and nutrients,but also good light intensity within the watercolumn, because many primary producers arephototrophic (such as green and red algae) orare mixotrophic (many corals and large benthic

ESC10 9/5/06 3:09 PM Page 355

356 PIERS LARCOMBE

Foraminifera). Other major producers of bio-genic sediment are molluscs and bryozoans,which feed on organic matter produced withinthe water column by phytoplankton, and henceultimately also rely on light.

Given the above, shelves dominated by bio-genic sediment production generally have lowrates of terrigenous input, because light inten-sity is decreased by high rates of sediment inputand/or reworking by shelf processes. Globally,major regions of tropical carbonate sedimenta-tion include the seas off Indonesia, northernWestern Australia, the Bahamas and the ArabianGulf (Chapter 9), and there are some areas ofcarbonate production in cooler waters, mostnotably the Lacepede shelf, southern Australia(James et al. 1992) but also in Brazil (Gischler &Lomando 1999; Testa & Bosence 1999). Thereare mechanisms for relatively rapid breakdownof some biogenic sediments, through bioerosionand solution, so that sediment production ratesneed to be distinguished carefully from eventualsediment accumulation rates (Buddemeier et al.1974).

Many organisms in these environments arehighly environmentally specific, and thus have

potential to form part of assessments of the‘health’ of modern ecosystems, or are potentialpalaeoenvironmental indicators. Notable in thisregard are foraminifers and diatoms, but otherrelevant fossils include dinoflagellate cysts, spongespicules, pollen, seeds, charcoal and ostracods.There is, however, a general underutilization ofmicrofossil-based techniques in studies of mod-ern environmental sedimentology. As an exam-ple, the broad use of diatoms in environmentalscience is well recognized (Stoermer & Smol1999) and they are particularly useful in studiesof environmental pollution of freshwater lakesand streams. Even in estuarine environments,however, there are relatively few applied studiesusing diatoms (Sullivan 1999), and there are evenfewer studies on continental shelves. Foramini-fera are increasingly used in applied environ-mental studies (Scott et al. 2001), but not yet foropen-shelf environments. There appears to beinsufficient work specifically designed to deter-mine the tolerances of benthic marine micro-organisms to various forms of pollution (Sullivan1999). The environmental constraints upon manylarger organisms are also poorly documented.For corals, there are some generally well-known

142

444

131

30

1438

150

18

28

154

256

210

42

17

113

232

290

67

110

31 133

56

59 25

2863228

3000

1738930

100

Sediment Yield(million tonnes per year)

Fig. 10.4 Volume of riverine silt supplied to continental shelves. (Adapted from Milliman & Meade 1993.)

ESC10 9/5/06 3:09 PM Page 356


environmental ranges for their occurrence (e.g.Potts & Jacobs 2003; Chapter 9), but these rangescover a variety of corals, some of which havevariable ways of obtaining energy and some ofwhich are adapted for variable conditions.

10.2.3 Sediment accumulation processes anddisturbance events

The wide distribution of modern continentalshelves means that, as a group, they experience a range of environmental processes (Fig. 10.3)and thus contain a diverse group of depositionalenvironments. The energy required to trans-port sediments can be derived from a number of sources. In most cases, the dominant energysupply is either tidal or weather-related, al-though there is generally a mixture of both. It isalso important to consider the relative impact ofdaily (generally low-energy) versus episodic high-energy phenomena (e.g. hurricanes, tsunami).Finally, there are very well-developed models of across-shelf transport (e.g. Fig. 10.5), but it isnow acknowledged that, for most shelves, thealong-shelf component of sediment transport isgreater than that across-shelf, especially wherethe coastline (or shallow bathymetry) is relativelystraight (e.g. Great Barrier Reef shelf, Otagoshelf, Texas–Louisiana shelf, this chapter).

10.2.3.1 Tide-influenced shelves

Tidal forces affect the whole globe, but their influ-ence on the world’s oceans is heavily modifiedby the bathymetry of the continental shelves andby the shape of the coastlines. Tides may occurdaily (diurnal) or twice-daily (semi-diurnal): formost places around the world, the semi-diurnaltide is dominant. The dynamics of many contin-ental shelves are dominated by tides, especiallywhere tidal ranges are high (Chapter 1), but evenin areas of low tidal range, tidal currents can bestrong locally, for example in the Torres Strait,Australia (Harris 1989, 1991) and the HayasuiStrait, Japan (Mogi 1979). On most shelves, the flood and ebb current directions are gener-ally opposed, but as they change in speed theyalso change direction, so that, overall, the tidedescribes an open ellipse, with a net direction andmagnitude (Stride 1982). Should the thresholdfor bed sediment transport be exceeded by oneor both of the tidal currents, net bed sedimenttransport results. Over long periods, the result is a pattern of regional bed-sediment-transportpathways, which begin at bed sediment ‘partingzones’, where they tend to be associated with lagsurfaces, and end at bed sediment ‘convergences’,where there is an accumulation of sandy sedi-ment. On the UK shelf, individual transport

Estuarine plume

Estuary

Dep

th (m

) 10

20

30

40

Waves

Beach

Wind-driven along-shelf flows

Rips

Internal waves

Physical andbiological mixing

of sediment column

Steep gradients incross-shelf sediment flux

Bottom boundary layer

Infragravity waves and shear waves

Resuspension by storms

Deposition and consolidation

Bioturbation

Inner shelf Mid-shelf

Verticaldiffusion

Fig. 10.5 Processes of cross-shelf transport. (Adapted from Nittrouer & Wright 1994.)

ESC10 9/5/06 3:09 PM Page 357

358 PIERS LARCOMBE

pathways are up to 550 km long and 170 kmwide (Belderson et al. 1982). The suite of sedi-mentary facies and bedforms present along suchpaths is well-defined on tidally dominated shelvesand is related both to the peak speed of the tidalcurrent and to sediment availability.

Volumetrically, there is generally little prac-tical impact that humans can have on these sediment transport pathways. A small transportpathway of 30 km in length, 5 km wide and 0.5 m average thickness would contain around130 × 106 t of sediment (assuming porosity of30% and density of 2.5), which is, for example,around five times the total weight (25 × 106 t) ofaggregate extracted from UK waters each year.Although apparently reassuring, this is an over-simplified view because sediment transport ratesalong the pathways are generally poorly known,and thefefore there are difficulties in assessing asustainable rate of removal of material. If therate of removal exceeds the rate of resupply, orthe deposits are relict, there is the potential tocreate a long-term change in bed habitats.

10.2.3.2 Storm-influenced shelves

Other reviews have noted that most (perhaps80%) of the world’s shelves are storm-dominated(e.g. Johnson & Baldwin 1996; Allen 1997). Theeffects of both waves and wind-driven currentsare relevant. Low-pressure weather systems overthe sea generate waves with a range of periodsand heights, which may influence shelf sedimenta-tion directly. Sedimentation on some shelves,however, can be dominated by swell waves gen-erated by storms in adjacent sea areas. In deepwater, waves travel at speeds relative to theirwavelength,

C = √(gλ /2π)

i.e. wave speed = √[(gravity × wavelength)/(2 × π)]

so that long waves travel fastest away from thecentre of a storm. Long waves also carry themost energy and transfer that energy deepestinto the water column, so that they are able tomobilize sediment on relatively deep areas ofcontinental shelves. As well as generating waves,

wind blowing over the sea surface drags wateralong with it, forming wind-driven currents.These currents are therefore fastest at, and oftenlimited to, the sea-surface, but where the wind is of sufficient strength, duration and fetch, thewhole depth of the shelf water mass may beinfluenced. A general rule-of-thumb is that thespeed of the surface wind-driven current is up to c. 3% of the wind speed (as measured 10 mabove the sea surface). Therefore, during storms,wind-driven currents can be of significant mag-nitude (see also Case Study 9.2).

It is useful to distinguish storm-driven fastunidirectional flows combined with waves,compared with the effects of long-period wavesacting alone. In this context, storms include hur-ricanes, typhoons and cyclones, which directlyinfluence the shelf with wind-driven currents,waves and land runoff. The Caribbean regionexperiences an average of seven hurricanes each year, some of which enter the Gulf ofMexico. The Texas–Louisiana shelf has an evenslope of around 1:1000, with the 80 m contourlocated c. 80 km offshore. In September 1961,Hurricane Carla, category 5, took 3 days to pass 400 km north-north-west across the shelf,producing a storm sand bed that extended for > 200 km along the shelf. The bed was preservedmostly within the 30 m contour, was thickest, at up to 9 cm, within the 20 m contour and wasrecognizable down to the 50 m depth contour.At the surface, the bed decreased in grain sizeoffshore. The storm sand bed fined upwards,from a sharp basal surface, through planar lam-inations to shallowly inclined laminations up to a gradational upper surface. Close to shore,the upper surface was sharp and truncated(Snedden & Nummedal 1991). These sediment-ary data are consistent with numerical modelsof shelf flow, which predicted that the windregime would have produced:1 a shoreward flow to the north-west at the surface;2 a fast shelf-parallel flow to the south-west of> 1.5 m s−1 in depths around 20 m;3 an oblique offshore return flow to the south-south-west or south near the bed (Forristall et al. 1977; Keen & Slingerland 1993).

ESC10 9/5/06 3:09 PM Page 358


Further, the oscillatory action of waves wouldhave dominated in shallow water. Hence, fastflows would have mobilized the sands and siltsof the inner shelf south-west along the shelf, andthe oblique offshore currents would have trans-ported sediments offshore, with finer sedimentsdeposited in deeper water.

The seawards-fining and seawards-thinningstorm bed described above, however, is not auniversal feature. Shelves of different morpho-logies and sedimentary regimes can have dis-tinctively different storm sediment dynamics onphysical scales of tens to hundreds of kilometres.For the Great Barrier Reef (GBR) shelf, the stormbed coarsens and thickens seawards (Gagan et al.

1988, 1990; Larcombe & Carter 2004). Further,and also in marked contrast with the Texas–Louisiana shelf, the GBR storm bed is best preserved on the inner shelf, where it is buriedbeneath the depth of subsequent bioturbation byimmediate post-storm sedimentation. Storm-bedpreservation is a function of cyclone recurrenceinterval (Case Study 10.1), storm-bed thicknessand the rate and depth of between-cyclone bio-turbation (Gagan et al. 1988, 1990). In passing,it has long been recognized that the frequentpassage of storms is not incompatible with thepresence of abundant and apparently delicatebenthic organisms on and in shelf sediments(e.g. Vaughan et al. 1987).

Case study 10.1 Time-scales of cyclone-driven sedimentary processes on the northern Great BarrierReef shelf

Like most regions of the Great Barrier Reef shelf, the shelf off Cairns (latitude 16°50′S) can bedivided into three sedimentary regions(Case Fig. 10.1A), which in this region consist of:1 the inner-shelf sediment prism (at depths of 0–20 m and out to 10–15 km from the coast)formed of bioturbated muddy sand;2 the middle shelf (depths of 20–40 m), narrow at this point at only 12 km wide, which ismostly formed of bioturbated mixed quartz and calcareous gravelly muddy sands;3 the outer-shelf reef complex, formed of a series of steep-sided patch reefs between which areshelly calcareous sandy gravels at depths of 40–80 m.

The climate is strongly seasonal, with a summer-monsoonal climate, reflected in the shelfhydrodynamic regime. In summer months (November–March), the weather is punctuated byoccasional cyclones, during which shelf waters are influenced by the discharge of muddy riverplumes and by wind-driven strong northward along-shelf flows (see also Case Study 9.2).

Major sediment transport events on the shelf are thus generally controlled by cyclones (CaseTable 10.1), the most severe (and least frequent) of which generate most sediment transportand may rework the combined results of all intervening (and less severe) cyclones, so that theresulting sedimentary record is dominated by a few cyclone beds. The stratigraphical record of cyclones is varied in nature, patchy and spasmodic, and mostly only available for the periodsince c. 5.5 kyr BP, the mid-Holocene sea-level highstand. Beach ridges on the Cairns coastalplain have an average interval of c. 280 years between the emplacement of successive ridges(Jones 1985). Equivalent figures for chenier ridges to the north and south are 177–280 years(Nott & Hayne 2001; Case Fig. 10.1B).

On the Cairns inner shelf, nearshore sediment cores display one or more sharp-based, finingupward, shell hash and sand to mud beds, interpreted as representing storm deposits (Carter et al. 2002). Radiocarbon dates from shells indicate ages of 3100, 2980 and 2830 yr BP for the three beds, indicating > 100-year periods (120 yr and 150 yr) between successive majorcyclones. On the inner shelf south of Cairns, modern storm beds (Cyclone Winifred, 1986,

ESC10 9/5/06 3:09 PM Page 359

360 PIERS LARCOMBE

Barron River

16o45' S

17o00' S

Trinity Bay

Batt Reef

ScottReef

NormanReefSaxon

Reef HastingsReef

ArlingtonReef

GreenIsland

FitzroyIsland

ThetfordReef Moore

Reef

Elford Reef

Sudbury Reef

Cairns

N

0 (km) 20

16o30' S

Trinity

Opening

Grafton

Passage

10 m

30 m

100 m

Innershelf

Middleshelf

Outershelf

Chenier plain

Core

Core

Case Fig. 10.1A Bathymetry of the shelf near Cairns, showing inner shelf, location of cheniers, core sites on inner and mid-shelf, and the outer shelf reef complex.

Case Table 10.1 The main sedimentary effects of cyclones across the main shelf environments of the GBR. (Sources: Hopley 1982; Gagan et al. 1988; Nott & Hayne 2001; Larcombe & Carter 2004.)

Shelf sedimentary environment

Coastline or island perimeter

Inner shelf

Middle shelf

Outer shelf

Main sedimentary features

Beach erosion, formation of chenierridges at muddy coastlines or beachridges at sandy coastlines

Unmix sediment, transport it along shelf,creating large dunes of shelly gravel and agraded bed, receive sediment from riversand from erosion on the middle shelf

As above plus formation of sand/gravelribbons

As middle shelf

Indicated average storm recurrence interval (yr)

177–280 (cheniers)

120–150 (shellbeds)

360–7600 (shellbedsat 40 m depth)

Sediment accumulation or erosion?

Variable

Net accumulation

Net erosion. Sedimentsexported both landward and seaward

Net sediment accumulation

ESC10 9/5/06 3:09 PM Page 360


category 3) comprise a moderately well sorted, graded bed of terrigenous sand–mud, locallywith a basal shell lag, and 5–20 cm in thickness (Gagan et al. 1988), which was remixed by bioturbation within 12 months. The c. 4 m thick Holocene inner-shelf sediment prism atCairns thus represents the stratigraphical equivalent of about 40 category 3 cyclones, implyinga recurrence interval of c. 140 yr.

On the Cairns middle shelf, a series of five radiocarbon dates from a core located in 40 m of water near Green Island, shows several storm beds in the past 1800 years preserved at anaverage interval of 360 years, and in nearby cores average intervals are up to 600 years, indicat-ing the generally erosive nature of the middle shelf ‘cyclone corridor’. Thus, stratigraphicalinformation from shelf cores and chenier ridges indicates an average return interval for majorcyclones (category 3 or higher) of between 150 and 300 yr. Although recurrence intervals and magnitudes will be different, similar processes and products are likely to occur on othertropical storm-influenced shelves, and perhaps other ‘rimmed’ shelves such as Belize and NewCaledonia (section 10.2.3.6), although there are relatively few data on such issues to date. The long recurrence intervals and high magnitudes of such disturbances are important con-siderations for managers of shelf environments, such as the Great Barrier Reef Marine ParkAuthority, who in their zoning scheme define allowable human use of different parts of the shelf(section 10.5.2.1).

Relevant reading

Carter, R.M., Larcombe, P., Liu, K., et al. (2002) The Environmental Sedimentology of Trinity Bay, far NorthQueensland. Final Report, James Cook University and Cairns Port Authority, 97 pp.

Case Fig. 10.1B Dated chenier-ridge sequences after Nott & Hayne (2001). Study sites and storm deposit data: (a) locationmap of study sites; (b) stratigraphical relationship of storm deposit/ridges on Curacoa Island (top) and Princess Charlotte Bay(bottom). Successive storm deposits are numbered accordingly. Mean reservoir-corrected radiocarbon age (in yr BP) for eachridge is shown above traces. Note progressive increase in age with distance inland. AHD is Australian Height Datum.

Gulf of Carpentaria

Great Barrier Reef

Princess CharlotteBay

CairnsCuracoa

Island

North-eastAustralia

300 km

54321

3

2

1

0

Met

res

AH

DM

etre

s A

HD

0 400 800

0 200 400

Distance from shore (m)

Supratidal mudflats

12

3 & 4

5

6

78

9 10

11

12

13

1415

16

17

18

1920

21 22

35 85 120

530

1010

1100

1475

1675

2065

2240

2540

2830

3025

3360

3640

3780

4070

4230

4410

4820

5095

5620

1

2

3 4 5 67

8

9 1011

12

70 300

465

566

785

1060

1190

1360

1730

1930

1965

2525

(a) (b)

ESC10 9/5/06 3:09 PM Page 361

362 PIERS LARCOMBE

10.2.3.3 Wave-influenced shelves

The Southern Ocean generates numerous power-ful storms, and shelves to its north are subject tothe impact of swell waves. On the south-facingLacepede shelf, South Australia, sedimentaryfacies are generally arranged in zones parallel tothe coast and mud is deposited only below depthsof 140 m (James et al. 1992). The sedimentaryregime of the South Otago shelf (Fig. 10.6) locatedon the south-east of South Island, New Zealand,can be characterized as being dominated by the action of swell waves plus storms (Carter et al. 1985; Carter & Carter 1986). Here, swellwaves typically mobilize the sandy sea-bed downto depths of 30 m, but a combination of theregional Southland Current, locally strong tides,plus storms, is required to mobilize fine sand todepths of c. 75 m. The grain size of the modernterrigenous inner-shelf sand wedge, which com-menced deposition in the past 6500 yr, fines seawards and northwards, with higher mud andbiogenic content at its seawards edge. A complexmineralogy and suite of heavy minerals, com-bined with textural trends and its overall mor-phology clearly indicate net northward sedimenttransport. On the inner shelf, the northern frontof the sand wedge has advanced northwards dur-ing the Holocene highstand at an average rate

of 30 m yr−1, and the sand is formed into largedunes in places. To seawards, at depths beneathc. 60 m, lies a gravel and sand facies related to early Holocene lower sea-level, the southernsandy portion of which is being slowly reworkedby modern storm-driven processes into sand ribbons, aligned along the shelf.

10.2.3.4 River-influenced shelves

Some continental shelves have their oceano-graphic and sedimentary regimes heavilyinfluenced by river input, in the form of largevolumes of freshwater and/or very high rates of sediment input (e.g. the Yellow Sea–EastChina Sea, Amazon). With such large systems,riverine, estuarine and shelf processes mergeacross the shelf and, with such muddy systems,fluid mud processes close to the sea bed becomeimportant (Kineke et al. 1996) along with fac-tors such as sediment oxygen concentration,organic content and bioturbation (Johnson &Baldwin 1996).

The Amazon shelf receives c. 1.2 × 109 t yr−1

of land-derived sediment from the Amazon River(Milliman & Meade 1983; Dunne et al. 1988;Allen 1997), which is distributed more than 100 km across and 400 km northwards alongthe shelf, and down to depths of 70 m. Sediment

Gagan, M.K., Johnson, D.P. & Carter, R.M. (1988) The Cyclone Winifred storm bed, central Great Barrier Reefshelf, Australia. Journal of Sedimentary Petrology 58, 845–56.

Gagan, M.K., Chivas, A.R. & Herczeg, A.L. (1990) Shelf-wide erosion, deposition, and suspended sedimenttransport during Cyclone Winifred, central Great Barrier Reef, Australia. Journal of Sedimentary Petrology60, 456–70.

Forristall, G.Z., Hamilton, R.C. & Cardone, V.J. (1977) Continental shelf currents in Tropical Storm Delia:observations and theory. Journal of Physical Oceanography 7, 532–46.

Hopley, D. (1982) The Geomorphology of the Great Barrier Reef: Quaternary Development of Coral Reefs.Wiley-Interscience, New York.

Hubbard, D.K. (1992) Hurricane-induced sediment transport in open-shelf tropical systems – an example fromSt. Croix, U.S. Virgin Islands. Journal of Sedimentary Petrology 62, 949–60.

Jones, M.R. (1985) Quaternary Geology and Coastline Evolution of Trinity Bay, North Queensland.Publication 386, Geological Survey of Queensland, Brisbane, 27 pp.

Larcombe, P. & Carter, R.M. (2004) Cyclone pumping, sediment partitioning and the development of the GreatBarrier Reef shelf system: a review. Quaternary Science Reviews 23, 107–35.

Nott, J. & Hayne, M. (2001) High frequency of ‘super-cyclones’ along the Great Barrier Reef over the past 5000 years. Nature 413, 508–12.

ESC10 9/5/06 3:09 PM Page 362


dispersal on this highly dynamic shelf is influencedby the inertia of the river flow, the along-shelfGuiana Current, which flows at 35–75 cm s−1

to the north-west, waves driven by trade-windsand by locally strong tidal currents (see alsoCase Study 9.1). The sediments are sands atriver mouths, and further along the transportpath are clayey silts, then interbedded silts andsands, and sandy silts, which form the top of thedelta and occupy the zone of highest oceano-graphic energy. Immediately to seawards, indepths of 40–60 m, is a zone of faintly laminatedmud, where the highest sediment accumulationrates are located, and the most abundant benthicbiology. Over time-scales of 100 years or so,these accumulation rates may reach 10 mm yr−1

(Fig. 10.7). Down-core sulphur concentrationsand 210Pb profiles indicate catastrophic accumu-lation events of up to 5 m yr−1, as well as epi-sodes of high rates of reworking and sedimentremoval (Aller & Blair 1996; Allison et al. 1996;Sommerfield et al. 1996; see also Kineke et al.1996). At their seaward edge, these Amazon-derived sediments become mottled and biotur-bated, and beyond is a zone of relict fine sands.Organic-rich laminae may occur widely acrossthe shelf, derived from either seasonal riverinput and/or plankton blooms. A large area ofthe shelf sediments (31,000 km2) is gas-enriched(Figueiredo et al. 1996), with the gas formedfrom breakdown of organic matter derived frommarine and terrestrial sources.

Terrigenous sediments

Modern sand

Modern mud

Relict gravel

Relict/palimpest sand

BiogenicRelict/palimpest sand and gravelwith local modern deposits

Bryozoan meadows

Shelf processesDirection of inner shelf transport(wave and storm dominated)Direction of mid- to outer-shelf transport(Southland Current and storm dominated)

Tide-dominated shelf

Sand wave field

Sand ribbons

TaieriRiver

TokomairiroRiver

CluthaRiver

800 m

Fig. 10.6 Thesedimentary facies andtransport regimes of theSouth Otago shelf.(Adapted from Carter et al. 1985.)

ESC10 9/5/06 3:09 PM Page 363

364 PIERS LARCOMBE

10.2.3.5 Oceanic current-influenced shelves

Although sedimentation on many shelves is influ-enced to some degree by oceanic currents, suchas on the Amazon, Newfoundland (Barrie &Collins 1989) and outer Saharan shelves (Newtonet al. 1973), several cases occur where the intrud-ing oceanic current (generally a western boundarycurrent) is a major contributor to sediment dis-persal, especially on the outer shelf, and in placeson the mid-shelf (Case Study 10.2). The longNW–SE–orientated shelf of north-east SouthAmerica is influenced by the Guiana Current,which flows at 35–75 cm s−1 to the north-west,and which supplements strong tidal currentsand waves. To the extreme east is the north-eastBrazilian shelf, which is relatively starved of land-derived sediments, but which contains a suite of coast-parallel zones of large-scale carbonateand siliciclastic bedforms, driven by oceanic and wind-driven flows, with superposed smaller

scale bedforms generated by waves and tides(Testa & Bosence 1999).

10.2.3.6 Wind-driven-current influenced shelves

Except during storms, wind-driven currents aregenerally only weak, but they may be sedimento-logically significant because they may carry finesediment placed into suspension by tides or fair-weather waves. On the Great Barrier Reefshelf, during fair-weather conditions, trade-windsblow from the south-east, along the shelf. Shelfsediment transport results primarily from wave-induced resuspension, combined with north-ward-directed wind-driven currents and coastallongshore drift (Belperio 1983, 1988; Larcombeet al. 1995; Orpin et al. 1999). Suspended sedi-ment concentrations of 10–100 mg L−1, causedby resuspension of sea-bed mud, occur in a 2–10 km wide coastal belt, which drifts north at rates of up to 15 cm s−1 under the influence of

52o W 51o 50o 49o 48o 47o 46o6o

5o

4o

3o

2o

1o

0o

1o

2oS

100 m

5 m

0 100 km

Marajo Island

52o W 51o 50o 49o 48o 47o 46o6o

5o

4o

3o

2o

1o

0o

1o

2oS

100 m

5 m

Amazon River

0 100 km

Marajo Island

Physically stratified sand

Interbedded mud/sand

Proximal-shell sandy silt

Faintly laminated mud

Mottled mud

Organic-rich laminae

Amazon River

0

0.11

2

35

10

Sediment accumulation rate mm yr −10 m

6 m

0 m

3 m

0 m

9 m

0m1 m

0 m

5 m

0 m0.5 m

Fig. 10.7 Sedimentary facies and accumulation rates on the continental shelf off the Amazon River. (Adapted from Kuehl et al. 1986.)

ESC10 9/5/06 3:09 PM Page 364


Case study 10.2 The south-east Australian and south-east African continental margins: shelfsedimentation and dispersal influenced by intruding oceanic currents

Off southern Queensland, Australia, shelf sedimentary facies include inner-shelf quartz sands(0–50 m depth), mid-shelf mixed carbonate–quartz sands (40–80 m depth) and, at a range ofdepths, but mostly on the outer shelf (depths > 40 m), pebble- to cobble-sized rhodoliths (Harriset al. 1996). Rhodoliths are a type of coated grain, with a laminated internal structure related toepisodes of growth of coralline algae. With repeated movements, the laminations formed are con-centric, and the threshold for movement of 50% of the rhodoliths in this area is 45–80 cm s−1,depending on their size. Current-meter data indicates that the southward-flowing East AustralianCurrent (EAC) episodically intrudes onto the shelf, moderating the tidal flows and inducingnear-bed current speeds of up to 130 cm s−1, even at depths of over 70 m (Case Fig. 10.2A). Fordepths of c. 80 m, only the largest storm waves can create oscillatory currents at the bed of over 60 cm s−1, and tidal currents alone are relatively weak (< 20 cm s−1). The EAC adds up to

Case Fig. 10.2A (a) The outer part of the southern Queensland shelf andassociated currents, and current-metermoorings located off northern FraserIsland. (b) Hourly average current speeds and direction on the outer part of the southern Queensland shelf.Note the residual currents to the south-south-east, indicating thepresence of the East Australian Current.(Adapted from Harris et al. 1996.)

East

Australian

Current

Study area

Australia

Great B

arrier Reef

FraserIsland

153o E

25o S

26o

20 40 60

120

140

20

0 (km) 20

1 23 4

5 67

Current metersite

1

50 cm s−1

N

Current meter 2

10 Days 25 302050

Current meter 4

(a)

(b)

ESC10 9/5/06 3:09 PM Page 365

366 PIERS LARCOMBE

55 cm s−1 to the speed of near-bed currents, initiating movement of the rhodoliths and thusforming a major control upon the presence of this facies on this shelf.

A particularly well-documented example of shelf sediment transport dominated by anintruding oceanic current is the south-east African shelf (Flemming 1978, 1980, 1981, 1988)(Case Fig. 10.2B). Here, the western boundary current is the Agulhas Current, which flows tothe south-south-west, parallel to the south-east African coast, at speeds up to 2.5 m s−1. Suchspeeds easily entrain sand and gravel on the sea bed and induce the development of large sedimentary bedforms such as large and very large transverse dunes (sensu Ashley et al. 1990)in symmetrical, asymmetrical and ‘cat-back’ forms (i.e. those where the crest is reversed in orientation compared with the overall bedform).

There is also a severe swell-wave regime, with regular large waves of long period. Given the microtidal regime (i.e. maximum tidal ranges < 2 m), the impact of the Agulhas Current isthe dominant mechanism driving sediment transport pathways. At major offsets in the con-tinental margin, the Agulhas Current overshoots the edge of the margin, and, at some distancedownstream, the flow ‘reattaches’ to the margin (Darbyshire 1972). Flow divides at these reattachment zones, so that some sediments to its north are transported back towards the offsetin a major eddy, and to its south are transported along the margin. As might be expected, thesezones tend to be relatively coarse-grained, ranging from fine sand up to gravel lag pavements.Long-term fluctuations in the current lead to changes in the location of the bedload partingzones of 10–100 km, so that in places a complex suite of sedimentary bedforms occurs whichindicates that bedload sediment transport reverses its direction. The suite of sedimentary faciesand bedforms on the shelf is very similar to that observed on many tidally dominated shelves,such as the UK shelf (section 10.2.3.1).

Africa

Maputo

Durban

Walvis Bay

CapeTown

0 (km) 800

Study area

Agul

has

Cur

rent

Moz

ambi

que

Cur

rent

Coastal upw

elling

BenguelaCurrent

ReturnAgulhasCurrent

Subtropical Convergence

−20 m−40 m

−60 m−80 m

−200 m

Agulh

as C

urre

nt0

510

Distance offshore (km)

Inner shelfTerrigenous facies(wave dominated)

Outer shelfRelict carbonatefacies(current dominated)100

80

60

40

20

0Alongsh

ore distance

(km)

222 x

22 x

1 xScale of distortion

Modern sand prism

Relict gravels

Underwater dunes

Sand ribbons

Direction of littoral drift

Direction of dunesand transport

Case Fig. 10.2B Location of Agulhas shelf with regional ocean current patterns and detail of the shelf-parallel sedimentaryfacies. (Adapted from Flemming 1980, 1981.)

ESC10 9/5/06 3:09 PM Page 366


the wind-driven, northward, along-shelf current.Where the coastline is relatively straight, themuddy sediment is transferred northward alongthe shelf, but where it is indented, fine sedimentsbecome trapped in inner-shelf embayments behindheadlands or sand spits.

Trade winds also dominate sedimentation onthe western New Caledonian shelf, in the south-west Pacific Ocean. Here, the shelf lagoon is 100 km in length, averages c. 18 m in depth,and has a funnel shape, narrowing from 40 kmwide in the south-east to about 5 km in the north-west, constrained to seawards by a series of longbarrier reefs. The maximum tidal range is c. 1.7m, and maximum tidal currents are c. 0.2 m s−1.The prevailing winds in the south-west lagoonare moderate to strong south-east trade winds(c. 18 knots) that blow for more than 200 days a year (November–May), and which drive shelfcurrents north-west along the lagoon (Douilletet al. 2001). Wind-driven currents dominate theoceanography when wind speeds are > 2 m s−1

(Douillet 1998). Although data are relatively few,resuspension and resedimentation can accountfor more than 80% of total sedimentation (Clavieret al. 1995), and wind-driven currents assist the shelf to remain relatively free from muddy

sediments, confining the accumulation of muddymaterial to embayments in the complex coast-line (Debenay 1987).

10.2.3.7 Additional shelf sedimentary processes

As noted above, sedimentary processes on muddyshelves are influenced by factors such as dissolvedoxygen concentration, sediment organic matterand bioturbation. On other shelves, especiallywhere carbonate producers are the primary sedi-ment source, various oceanographic factors asso-ciated with the water column (nutrient supply,water clarity, water temperature) are all import-ant factors (Reading 1996). Regarding sedimen-tary ‘disturbance’, natural hydrocarbon seeps can be important on some continental shelves,because pockmarks produced by escaping fluidscan be 0.5 to 20 m deep and 1 m to 1 km indiameter (Hovland & Judd 1988). Pockmarksrepresent material released from the sea bed andmay contain an aggregation of biogenic materialin the centre (Dando 2001). Finally, in the lightof the (magnitude 9.0) earthquake of 26 Decem-ber 2004 beneath the sea bed off north-westIndonesia, which created a destructive tsunamithat killed more than 150,000 people, it is worth

Relevant reading

Ashley, G.M., Dalrymple, R.W., Elliott, T., et al. (1990) Classification of large-scale subaqueous bedforms: a new look at an old problem. Journal of Sedimentary Petrology 60, 160–72.

Belderson, R.H., Johnson, M.A. & Kenyon, N.H. (1982) Bedforms. In: Offshore Tidal Sands (Ed. A.H. Stride),pp. 27–57. Chapman and Hall, London.

Darbyshire, J. (1972) The effect of bottom topography on the Agulhas Current. Pure and Applied Geophysics101, 208–20.

Flemming, B.W. (1978) Underwater sand dunes along the southeast African continental margin – observationsand implications. Marine Geology 26, 177–98.

Flemming, B.W. (1980) Sand transport and bedform patterns on the continental shelf between Durban and PortElizabeth (southeast African continental margin. Sedimentary Geology 26, 179–205.

Flemming, B.W. (1981) Factors controlling shelf sediment dispersal along the southeast African continentalmargin. Marine Geology 42, 259–77.

Flemming, B.W. (1988) Pseudo-tidal sedimentation in a non-tidal shelf environment (southeast African con-tinental margin). In: Tide-influenced Sedimentary Environments and Facies (Eds P.L. De Boer, A. van Gelder& S.D. Nio), pp. 167–80. Reidel, Dordrecht.

Harris, P.T., Tsuji, Y., Marshall, J.F., et al. (1996) Sand and rhodolith-gravel entrainment on the mid- to outer-shelf under a western boundary current: Fraser Island continental shelf, eastern Australia. Marine Geology129, 313–30.

ESC10 9/5/06 3:09 PM Page 367

368 PIERS LARCOMBE

noting that the Pacific Ocean has an average ofone to two destructive tsunamis each year; i.e.they are geologically common. Although manycontinental shelves are clearly influenced bytsunami, there are few existing data from modernshelves on their flows or their deposits.

10.2.3.8 Inferring sedimentary processes fromsedimentary facies

Where direct measurements of shelf sediment-ary processes do not exist, the nature of long-term shelf sedimentation can be inferred fromthe characteristics of the deposits, using aspectsincluding the spatial distribution of sediment-ary facies and bedforms, gradients of sediment texture and composition, and the shallow stra-tigraphy. It is of course very helpful to date thesediments and hence infer the timing of the mainsedimentary events. Long-term patterns of sedi-ment dispersal sometimes can be demonstratedfrom regional variations in sedimentation. Onthe central inner-shelf of the Great Barrier Reef,Holocene sediment accumulation rates decreasewith distance from the main source, the BurdekinRiver, which discharges sediment onto the shelfat an average rate of 3–9 × 106 t yr−1 (Neil et al.2002). Holocene sediment accumulation ratesdecrease along the regional shelf-parallel sedimenttransport pathway, from c. 0.7 to < 0.1 mm yr−1

(Woolfe & Larcombe 1998) as a function of theregional trapping ability of a set of north-facingcoastal embayments. On the larger scale of theentire central and northern GBR shelf, the textureand composition of terrigenous sediments alongthe 10 m depth contour, although complex, alsoindicate general net northward shelf sedimentdispersal (Lambeck & Woolfe 2000).

Distance along a transport path sometimescan be reflected in terms of two sedimentary gradients:1 minerals of low resistance to abrasion andbreakage (e.g. carbonate grains) give way to moreresistant minerals (e.g. heavy minerals);2 grains become less angular and more rounded.Concepts of compositional and textural maturitythus result (Pettijohn et al. 1987), with immaturesediments tending to be close to source. Care is

needed applying such concepts to modern shelfsedimentary processes, because shelf sedimentscan reflect, to varying degrees, sedimentary pro-cesses that are no longer active, and there can bea variety of grain sources.

Increasingly, routine use of laser-diffraction andother high-resolution techniques indicates thatmany shelf sediments have polymodal grain-sizedistributions. Size modes, once identified, canreveal dispersal patterns. For example, on theCalifornian shelf, the distribution of the domin-ant size modes shows that Columbia River sedi-ments are transported north-north-west across theshelf, and decrease in size from fine sand to coarsesilt (Fig. 10.8). Very fine sands are held close tothe coast and accumulate at rates below c. 1.4 mmyr−1, whereas coarse silts dominate along a well-defined axis towards the north-north-west, rep-resenting the main transport path of suspendedsediment. Accumulation rates decrease alongthe main transport path from about 7 mm yr−1

to 3 mm yr−1. The sediment transport rate ofvolcanic ash across the shelf has been calculated,

> 4.2

2.8–4.2

1.4–2.8

< 1.4

Coast

Accumulation rates (mm yr−1)

Shelfedge

Astoria Canyon

0 20 km

Columbia River

Fig. 10.8 The Oregon–Washington shelf: accumulation ratesof sediments. (Adapted from Nittrouer et al. 1979.)

ESC10 9/5/06 3:09 PM Page 368


using the time of arrival of material derived fromthe Mount St Helen’s eruption of May 1980(Chapter 2), which had been carried to the shelfby the river. Only 17 months after the eruption,the ash had been transported to the north-north-west, at least 125 km along the shelf and 40 kmacross it (Ridge & Carson 1987).

Polymodal shelf sediments aid the delineationof shelf sedimentary facies and infer character-istics of the sediment environments. Woolfe et al.(2000) analysed the silt and sand fractions of 300samples from the central GBR shelf (Fig. 10.9)to show that the muddy sediments supplied bythe Herbert River (Group 2) either are trappedin sheltered environments close to the coast or aredeposited below the 10 m depth contour on opencoasts. Sediment accumulation rates measuredusing 210Pb and 137Cs profiles from core samplesare 0.7–12.3 kg m2 yr−1 (mean c. 4 kg m2 yr−1)(Brunskill et al. 2002). These muds are separatedfrom the coastline by an erosional (or at leasthighly mobile) sandy subtidal zone (Group 1),and to seawards, much of the shelf is occupiedby a muddy, medium to coarse-grained terrigen-ous or calcareous sand (Group 3) of very lowaccumulation rate (generally < 0.1 kg m2 yr−1).

Clearly, on many shelves, characterizing sedi-ments by mean grain size is a simplification. Vari-ous computer programs are used by industry topredict sediment transport on continental shelves,some of which use a single grain size. Althoughsuch a simple approach can sometimes fit thepurpose, predictive models of shelf sedimenttransport are increasingly being demanded inassessments of human impacts, whether regard-ing marine engineering projects or regional stud-ies of pollution dispersal and accumulation, sothat more complex models need designing andtesting. Porter-Smith et al. (2004) have com-bined information on shelf sediment texture withhydrodynamic models to delineate regions of theAustralian continental shelf, based on sedimenttransport processes.

10.2.3.9 Measuring sedimentary processes

Increasingly, studies on shelves use sea-bedmapping techniques (high-resolution swath-

bathymetry and digital side-scan sonar) to pro-duce three-dimensional digital terrain models of the sea bed, which show sediment distribu-tion and bedforms. Repeated surveys can shownet changes in bed elevation, and are now stand-ard industry practice in assessing sedimentaryimpacts of offshore activities. Grabs and corersremain essential elements of field studies (Alleret al. 2004), and nowadays include refined corersthat sample the water–sediment interface undis-turbed. Accumulation rates of sediments can bemade in muds using various radio-tracers (e.g.Brunskill et al. 2002).

Measurements of currents and waves are now generally of very high quality, for exam-ple using acoustic devices such as AcousticDoppler Current Profilers (ADCPs) deployedfrom buoys, vessels or at the sea bed. Instru-ment packages deployed at the sea bed canallow simultaneous measurements of near-bedsedimentary processes over periods of days toweeks (e.g. Cacchione et al. 1999). Sedimenttraps have limited use in shelf studies, because of strong horizontal transport and resuspen-sion events (e.g. Topçu & Brockman 2001;Thomas & Ridd 2004). At the bed itself, time-series measurement of sediment accumulationor erosion is possible at a single point (e.g.Larcombe et al. 1995; Thomas & Ridd 2004).Direct measures of bedform migration are pos-sible through use of video cameras, or sector-scanning sonar. Acoustic techniques also existto assess sediment transport, especially of sand(Thorne & Hanes 2002). Optical devices nowexist for measuring grain-size distributions in situ(Gartner et al. 2001; see also Van Walree et al.(2005) for progress on acoustic techniques), andalthough real-time field measurements of sedi-ment transport rates are increasingly common,they generally remain scientific and investigativetechniques rather than being applicable tools. The algorithms applied to satellite images areimproving (Binding et al. 2003), but are not yetsufficiently robust to be generally applicable tosedimentary studies. Most shelf sediment trans-port occurs near the sea-bed, where informationfrom satellites is weakest, especially when sedi-ment transport rates are high.

ESC10 9/5/06 3:09 PM Page 369

370 PIERS LARCOMBE

Herbert River

50

km

N

30

30

20

20

10

10

5

5

5

15

1520

Palm Islands

Cardwell

Australia

20

Lucinda

Island

StudyareaRockingham Bay

Hinchinbrook Island

Halifax Bay

(a)

(b)

Group 1

Group 2

Group 3

Group 4

Group 1 (n = 54)

Group 2 (n = 25)

Group 3 (n = 97)

Vol. %Group 4 (n = 32)

5 10 50 100 250 1000nm

45

30

15

0

45

30

15

0

45

30

15

0

45

30

15

0

(c)

Fig. 10.9 (a) The shelf off theHerbert River delta, central GreatBarrier Reef shelf, showingbathymetry (contours in metres) and sample locations (dots). (b) The four delineated sedimentgroups off the Herbert River delta and (c) their distribution. (Adapted from Woolfe et al. 2000.)

ESC10 9/5/06 3:09 PM Page 370


10.3 ANTHROPOGENIC ACTIVITIES – PROCESSES

AND IMPACTS

10.3.1 Anthropogenic inputs into shelf sediments

There is a broad range of sources of materialinputs into shelf waters, which includes rivers,the atmosphere, maritime transport operations(e.g. dredge material disposal), shipwrecks, off-shore hydrocarbon production facilities, smeltersand electricity generation plants. Most signifi-cant volumes of anthropogenic inputs into themarine environment occur within estuaries or atthe coastline, so few inputs occur beyond theinfluence of coastal processes. There are rela-tively few inputs of volumetric significance andhence immediate sedimentary consequence.

10.3.1.1 Offshore disposal of dredged material

Most dredging is performed in rivers, estuariesor coastal waters for navigational purposes, for underlying reasons of defence, trade, coastaldefence and/or recreation. Offshore, the mainsedimentary consequences of the emplacementof dredged material onto the sea bed are thatsea-bed habitats may be altered and that con-taminants may be introduced offshore. Thewebsites of the London Convention and OSPAR(Convention for the Protection of the MarineEnvironment of the North East Atlantic) pro-vide background information on the issue ofdisposal. Information on the total amount ofdredged material disposed of at sea is in-complete, but for Europe in 2003 it totalled 63 × 106 t dw (dry weight). Vivian & Murray(2002) have estimated that, world-wide, up to

about 1 × 109 t of dredged material might bedisposed of at sea.

10.3.1.2 Offshore dumping of sewage sludge

The dumping of sewage sludge at sea remains in use for some countries around the world,although sewage sludge dumping off the eastcoast of the USA ended in 1992, and in north-east Atlantic waters, it has been prohibited since1999. Contamination of the food web can occur by metals, organic and biosynthetic com-pounds, but in sedimentary terms, the organiccontent of emplaced material is high, break-down of which can lower oxygen concentra-tions in marine sediments and, in some cases,the overlying water as well. There is increasingeffort put towards reprocessing sewage sludgeon land. In some countries, pellets are being pro-duced for fertilizer, as fuel for industrial powerplants and for use in charcoal production.

10.3.1.3 Offshore hydrocarbon production facilities

Hydrocarbon extraction on continental shelvescan involve the discharge of contaminated sedi-ments onto the sea floor. In the North Sea, wherecontamination levels are typically very low insediments (Table 10.1), between 1 and 1.5 × 106 tof oil-based drill mud and rock cuttings are estimated to lie on the North Sea, derived fromthe rapid fall-out of discharged drill cuttingsonto the sea-bed below oil and gas productionplatforms. Such discharges are now prohibited,but in parts of the southern North Sea, past discharges resulted in numerous large, discrete,cuttings piles (Daan & Mulder 1996; CEFAS

Table 10.1 Summary of contaminant levels typically found in surface sediments from the North Sea (DTI 2001).

Location THC PAH PCB Ni Cu Zn Cd Hg (μg g−1) (μg g−1) (μg kg−1) (μg g) (μg g−1) (μg g−1) (μg g−1) (μg g−1)

Oil and gas installations 10–450 0.02–74.7 – 17.79 17.45 129.74 0.85 0.36Estuaries – 0.2–28 6.8–19.1 – – – – –Coast – – 2 – – – – –Offshore 17–120 0.2–2.7 < 1 9.5 3.96 20.87 0.43 0.16

THC, total hydrocarbon; PAH, polycyclic aromatic hydrocarbon; PCB, polychlorinated biphenyls.

ESC10 9/5/06 3:09 PM Page 371

372 PIERS LARCOMBE

2001; Department of Trade and Industry 2001),which are localized hotspots of contamination byhydrocarbons and a range of other compounds.Contamination tends to be limited to within a distance of about 500 m from most offshoreplatforms, but in some areas hydrocarbon-contaminated sediments extend to 2–8 km fromthe platforms (OSPAR Commission 2000). Inother areas of the North Sea, dispersal hasoccurred, preventing the formation of such cut-tings piles. Effects on benthic communities havebeen observed out to 3–5 km, although since cuttings discharges were prohibited, the zonesof impact have decreased.

10.3.2 Anthropogenic disturbances on shelfsedimentation

10.3.2.1 Dammed rivers

The practice of building dams on rivers to securewater supplies for humans or for hydroelectricpower plants has the general effect of decreasingsediment supply down rivers, particularly ofbedload, with potential effects on the coastaland nearshore zones (see also Case Study 8.3).On the open shelf there is a decreasing relative effect with distance down the transport path,because shelf processes influence the river-bornesediment supply. For the South Otago shelf,New Zealand (section 10.2.3.3), over the past9600 years, sediment input to the shelf has beenaround 2.1 × 106 t yr−1 of bedload, of whicharound half accumulates in the shelf-sandwedge and half is transported northwards toadjacent shelf regions (Carter 1986). Virtuallyall the 2.3 × 106 t yr−1 of suspended load receivedby the shelf is transported northwards along theshelf or offshore to the adjacent continentalslope. On one of the three main supplying catch-ments, the Roxburgh Dam built in 1961 on theClutha River traps 0.6 × 106 t yr−1 of bedload,which is around 30% of the total bedload supply to the entire South Otago shelf – this islikely to have affected sediment transport and/or the development of the Holocene inner-shelfsediment body. It is unknown whether thecoast-shelf sediment regime has adjusted to the

1961 dam, and the variable effect of oceano-graphic processes is unknown, given the relativelyhigh input of land-derived sediment. Regardingthe potential effects of dams on the South Otagoshelf-coast system, Carter (1986) considers thatthe key is whether the sediment being suppliedby a river has historically (i) become entrainedwithin the inner-shelf transport system, in whichcase there would be further impacts, or (ii) hastended to bypass the nearshore zone, in whichcase the effect would be to reduce sediment trans-port on the mid-shelf and reduce sand supply tothe adjacent shelf and coastline to the north.

10.3.2.2 Trawling

Globally, trawling may be the most intensive of anthropogenic disturbances to the sea bed.An estimated area equivalent to all the world’scontinental shelves is trawled every 2 years(Watling & Norse 1998), and in the North Sea,total trawling effort is equivalent to the wholeNorth Sea being trawled at least seven times a year. On the regional and local scale, how-ever, trawling is very patchy. As an example, the Dutch beam-trawl fleet visits some areas of the North Sea over 400 times per year butother areas not at all (Rijnsdorp et al. 1998;Trimmer et al. 2005), and there is also signific-ant within-year variation in overall trawlingeffort. Trawling may mix and resuspend sur-face sediments (those down to depths of a fewcentimetres), can disturb sediment down to adecimetre or more, can release nutrients into the water column and can influence the benthicbiology (Jennings & Kaiser 1998; Kaiser & De Groot 2000), particularly surface dwellers.A changed benthic biology can alter sedimentbiogeochemistry because of changes in bio-turbation and bio-irrigation. At some heavilytrawled sites in the North Sea, biogeochemicalprocesses in the upper layers of sediment, bothoxic and anoxic, may be unaffected by trawl-ing in the long-term (Trimmer et al. 2005), butin underlying anoxic sediments, mineralizationvia sulphate reduction may be stimulated by theextra disturbance, at least in areas where tidalenergy is weak.

ESC10 9/5/06 3:09 PM Page 372


Trawling activity is intense on the mid-AtlanticBight. Churchill (1989) studied the sedimentaryimpacts of trawling there, using regional data ontrawling activity, field observations and meas-urements of the changes in suspended sedimentconcentration (SSC) at 100–120 m depth. Over-all, trawling effort decreases with depth and dis-tance across the shelf (Fig. 10.10), but trawlingis concentrated in some areas, such as the ‘MudPatch’, south of Cape Cod, where some areas aretrawled completely over three or four times peryear. Seasonal variations in natural resuspensionand in trawling activity mean that trawling is asignificant generator of suspended sediment onthe shelf, except during winter and early spring.Once resuspended, material tends to movedownslope off the shelf, and Churchill (1989)calculated that 70,000 t yr−1 of sediment is lostdownslope from the ‘Mud Patch’, a loss equival-ent to 0.2 mm yr−1. This is a significant rate com-pared with the estimated rate of shelf sedimentaccumulation of 0.2–0.3 mm yr−1 (calculated by

Bothner et al. 1981a,b). Consequently, in places,trawling appears to be preventing shelf sedi-ment accumulation, as well as increasing ratesof sediment transfer off the continental shelf.

Further attempts to quantify the sediment-ary effects of trawling have been performed byPalanques et al. (2001) on the unfished muddyshelf of the north-west Mediterranean. The workincluded study of the physical characteristics ofthe bed sediments at 30–40 m depth, where thereis a natural nepheloid layer of SSC 2–3 mg L−1,which extends upwards from the bed for 3–8 m.In calm conditions, use of an Otter trawl with a4-m-wide line rope, towed along lines 2700 mlong, increased the SSC in the nepheloid layer to5 mg L−1. The impacts remain easily detectable5 days after trawling, when the total weight ofsuspended sediment in the area was 350 t, nearlythree times the original 120 t. Analysis of the siltcontent of sediments in core samples indicatedthat 2–3 cm of material had been eroded by thetrawl, and that 10% of this disturbed sedimentremained in the water column 4–5 days aftertrawling.

10.3.2.3 Aggregate dredging

Extraction of marine aggregates from continentalshelves is largely based on the need to obtaindeposits of a particular range of grain sizes (sandsand gravels), primarily for building and construc-tion purposes, but also for beach nourishmentand other minor uses. The volume of marineaggregates extracted from the continental shelvesaround Europe was 40 × 106 m3 yr−1 (for 1992–97), with the UK and The Netherlands by far themajor extractors (Table 10.2). For the UK, anaverage of 25 × 106 t yr−1 of sand and gravel isextracted from the shelf, representing about 20%of all sands and gravels used in the UK. Licencedextraction areas are concentrated on the southernand eastern shelf, close to the economic activityof southern England, although it is worth notingthat an area less than 12% of the total is actuallyused each year (Table 10.3).

Typically, aggregate dredging is performed by trailer suction hopper dredgers that operatewhile underway, leading to the production of

42o

41o

40o

39o

38o

37o

36o

75o76o 74o 73o 72o 71o 70o 69o

2000 m

0

0

0

0 0

0

0 0

000

0 0 0 0 0

0 0 0

0

0

0

1 1

1 1

32 13

69 5

19 12 3

33 19 29

3419

25

86 15 28 52

20 2 7 14 2

2 6 2231

19

93 8 19 1027 67 35 48 53 107

271413321

145140158

131

95

89755019

Area trawled/box area (%)U.S. trawlers 1985

60 m

44

76

Fig. 10.10 Distribution of trawling by USA vessels over themid-Atlantic Bight. The number in each box is the relative area ofsea bed trawled (i.e. [cumulative trawled area/box area] × 100).Data are for 1985. (Adapted from Churchill 1989.)

ESC10 9/5/06 3:09 PM Page 373

374 PIERS LARCOMBE

shallow linear furrows 1–3 m wide and 0.2–0.3 m deep (Kenny & Rees 1994, 1996). Otherextraction methods result in saucer-shaped de-pressions up to 200 m in diameter and 8–10 mdeep, with slopes of c. 5° (Wenban-Smith 2002;EMU 2004). Together with the direct physicalimpacts on the sea bed, typically 5% of dredgedmaterial is returned to the sea bed during thedredging process, but this may increase to 60%when screening is used to increase the propor-tion of gravel in the cargo. There are thus alsoimpacts from the settling of sediments throughthe water column back to the sea bed.

The impacts of aggregate dredging are de-pendent upon the nature of the substrate beingdredged. In physical terms, impacts include:1 The formation of dredge tracks or depressionson the sea bed (Fig. 10.11).2 Disruption of the surficial sediments, leadingto release of sands and finer sediments from thematrix and from underlying deposits.3 Generation of sediment plumes at the sea-bedby the action of the drag head, and plumes at thesea surface which then settle to the bed.

4 Changed grain-size characteristics of the bedsediments (McCauley et al. 1977), which mightbe an increase in gravel content through the ex-posure of coarser sediments (Kenny et al. 1998),an increase in the proportion of fine sands(Desprez 2000; van Dalfsen et al. 2000), loss of silt- and clay-sized grains (Boyd et al. 2003,2004) and organic material (EMU 2004).5 Temporary changes in the sediment trans-port regime at and near the bed. Typically, thisincludes the introduction of a local ‘slug’ of sandysediment onto the bed, derived from settling ofsediments rejected by the screening process onboard the dredge and from sands released fromthe sea bed. Initially, these sands are generallydeposited within a few hundred metres of thedredge site. Silts, clays and organic matter aredistributed more widely, and all are then liableto redistribution by waves and currents.

It has long been recognized that the sediment-ary environment (composition, texture, struc-ture and stability) is a major control upon thebenthic biological assemblage (Holme 1961,1966; Holme & Wilson 1985). Studies of impacts

Country Total × 103 (m3) Average × 103 (m3 yr−1)

Belgium 11,000 1833Denmark 30,500 5083France 13,200 2200Germany 17,000 2833Iceland No information No informationIreland Zero abstraction Zero abstractionThe Netherlands 104,200 17,366Norway 710 118Portugal No information No informationSpain No information No informationSweden Zero abstraction Zero abstractionUK 81,600 13,600

Table 10.3 Summary of UK extraction of terrestrial and marine aggregates (EMU 2004).

Total area Total area Percentage Volume extracted Maximum Average licensed (sand worked area worked annually (England size of site size of siteand gravel) annually annually and Wales)

Terrestrial 270 km2 150 km2 55 70 Μt yr−1 9.45 km2 < 1 km2

Marine 1300 km2 150 km2 11.5 23 Μt yr−1

Table 10.2 Volumes of marine sandand gravel extracted from the westernEuropean shelf for the period 1992–97(OSPAR Commission 2000).

ESC10 9/5/06 3:09 PM Page 374


of marine aggregate dredging on the sea bed havefocused on the benthic biological communities,although the primary sedimentary factors areincreasingly being recognized. In the UK, various‘biotopes’ are defined to help assess the natureof impacts, ‘recovery’ and mediation, these being(i) shallow-water mobile sands, (ii) shallow-waterstable gravels with transient sands and (iii) stable gravels (EMU 2004). Biological ‘recovery’requires the sediments to revert to pre-dredge or otherwise stable conditions, and tends to bemore rapid in mobile sediments such as sands inshallow waters (few months to 4 years). In con-trast, in stable areas (usually gravelly and deep),‘recovery’ of the fauna is much slower (up to > 15 yr) because of the presence of long-livedspecies (EMU 2004). There is great variation inthe nature and rates of faunal change ‘recovery’,however, and meaningful assessments of recoveryreally can be made only on a site-specific basis.

10.3.2.4 Marine mining

Marine mining differs from aggregate extrac-tion because mining exploits marine sedimentsspecifically for their mineral content rather thantheir texture. Such deposits fall into the categoryof ‘placer’ deposits, which contain economical

quantities of valuable minerals. Those mineralscontaining gold, diamonds, platinum and tin arethe most important. Such minerals tend to behighly resistant to water and abrasion, and ofhigh density (> 2.9). Physical concentration ofsuch minerals usually takes place as a result of transport, and on continental shelves theseminerals tend to be associated with deposits of high-energy environments that operated forextended periods.

Diamonds occur in sediments of the westernshelf of southern Africa. Exploration licences havebeen granted and active prospecting is takingplace in shelf areas off Namibia and South Africa,at depths down to 200–500 m. Off southernNamibia, diamond-bearing deposits are associ-ated with Pleistocene gravels that developed at the LGM coastline, near the modern 130 mdepth contour (Rogers & Li 2002). These beachgravels are now buried beneath a thin Holocenesequence of sediments (< 30 cm thick) associatedwith the Holocene progradation of the OrangeRiver delta. Overall, the thin Holocene sedi-ment package thus has a characteristic finingupward sequence representing the post-glaciallandward migration of sedimentary facies, over-lain by a coarsening upward progradationalmarine deltaic sequence, although bioturbation

A period immediately after dredgingA period of months after dredgingA period of years after dredging

A'

A

B'

B

A A' B B'

Ephemeralsand

Sand ribbon

Sandpatch

Sand sheet

Dredging lane

Licence area

Infill

Fig. 10.11 Cartoon of main sedimentary consequences of dredging. Hypothetical distribution of major sedimentary bedforms aroundan individual dredging lane. Net bed sediment transport is towards the left. A–A′ and B–B′ are sections across the dredging lane.

ESC10 9/5/06 3:09 PM Page 375

Gra

vel

Mud

San

d

100 90

80

70

60

50

40

30

20

10

0

0

10

20

30

40

50

60

70

80

90

100 0

1020

3040

5060

7080

9010

0

Gra

vel

Mud

San

d

100 90

80

70

60

50

40

30

20

10

0

0

10

20

30

40

50

60

70

80

90

100 0

1020

3040

5060

7080

9010

0

Are

a 1

- Jun

e 19

94A

rea

1 - F

ebru

ary

1995

A

rea

3 - J

une

1994

Are

a 3

- Feb

ruar

y 19

95

Text

ure

- ar

ea 2

Te

xtur

e -

area

3

50 40 30 20 10 0

%

-1

0

1

2

3

4

50 40 30 20 10 0

%

-1

0

1

2

3

4

50 40 30 20 10 0

%

-1

0

1

2

3

4

50 40 30 20 10 0

%

-1

0

1

2

3

4

Gra

b sa

mpl

esJu

ne 1

994

Gra

b sa

mpl

esFe

brua

ry 1

995

Gra

b sa

mpl

esJu

ne 1

994

Gra

b sa

mpl

esFe

brua

ry 1

995

Fig.

10.

12Se

dim

ent t

extu

re a

nd re

pres

enta

tive

grai

n-si

ze d

istri

butio

ns o

f the

sand

frac

tions

of s

edim

ents

from

the

mid

dle

shel

f off

Nam

ibia

: (le

ft) u

nmin

ed se

dim

ents

(u

nim

odal

ver

y fin

e sa

nds)

; (rig

ht) m

ined

sedi

men

ts (p

olym

odal

sand

s). (

Ada

pted

from

Rog

ers &

Li 2

002.

)

ESC10 9/5/06 3:09 PM Page 376


has mixed the thin sequence. Although sea-bedsediments deeper than 40 m are generally toodeep for common wave disturbance, and thuscontain silts and clays, sediments deeper than100 m are able to be mobilized, either by oscill-atory currents driven from long-period swellwaves, or by wind-driven currents. Over largeareas of the shelf, storms resuspend fine sedi-ments and may create transient nepheloid layersnear the bed. On the middle shelf, the naturalvery fine sands are thus very well sorted and uni-modal in nature.

On board the vessel, the mining processinvolves sorting of sediments into cobbles, pebbles and tailings (sand, silt and clay), whichfractions are then discarded over the side via aconveyor belt or a tailings pipe, so that miningactivity tends to increase the patchiness of thesea bed. Off southern Namibia, in mined areas,the well-sorted sands are replaced by poorlysorted and heterogeneous sediments, because of the addition of gravels and coarser sandsderived from the mined buried lowstand sedi-ments (Fig. 10.12). In areas of the inner shelf, indepths of < 40 m, storms have a relatively greatimpact in transporting gravel, compared withmining. Although immediate impacts tend to be local, the area mined is expanding rapidly,and ‘conservation corridors’ are being proposedbetween mining lanes to preserve the Holocenesequence in some areas, and to provide refugesfor benthic organisms to improve recolonizationof the substrate after the cessation of mining.Such zones are also proposed between aggregatedredging lanes (Boyd et al. 2004).

10.3.2.5 Offshore windfarms

For many countries, a major driver for thedevelopment of Offshore Renewable EnergyDevelopments (OREDs) is the attempt to reduceCO2 emissions in response to the Kyoto Pro-tocol and to diversify energy supply. The mainactive OREDs are offshore windfarms, whichcomprise an array of large individual wind tur-bines, each comprising an impellor (typically ofthree or four blades) located on top of a towerconnected to a monopile attached to the sea bed.

Between these units runs a network of cables,usually buried beneath sediment. The UK is aleader in the use of this technology, and late in2005, electricity was already being supplied tothe UK National Grid from three offshore wind-farms. The first sites each occupy c. 10 km2 ofsea bed and have up to 30 turbines per site, butthis is due to increase greatly in the near futureto as many as 200–300 turbines per site.

There are potential sedimentary effects duringthe construction, operation and decommission-ing phases of offshore windfarms (Rees, in Judd et al. 2003). In terms of physical sedimentaryeffects, applications for windfarm developmentsare considered on a site-by-site basis, requiring areview of the local and regional coastal processesand assessment of whether the structures mightchange sediment transport patterns, rates andpathways. Monitoring programmes are under-taken to test the predictions made by the Environ-mental Impact Statements. Such work buildsupon basic knowledge of sedimentary impactsof flows around structures (e.g. Allen 1984), andincludes accurate swath-bathymetric surveys(Fig. 10.13), repeated at intervals to determinesediment transport associated with bedformmigration, and overall changes in bed elevation.Of major interest are the processes of scouraround the monopiles. Observation and model-ling appear to indicate that, even in areas ofhighly dynamic sediment transport, the volumeof material removed by scour around individualmonopiles is insignificant compared with thatassociated with bedform migration across thewhole windfarm site. Further, spacing betweenmonopiles of 300–400 m appears to be sufficientto produce no combined significant sedimentaryimpacts (Fig. 10.13). This accords with the ‘ruleof thumb’ of little impact upon net flow at 6–10obstacle diameters downstream. Hence, althoughthere is a range of research aimed at improvingunderstanding of the environmental impacts ofwindfarms, including a wide range of ecologicaleffects (Gill 2005), there are unlikely to be majorsedimentary consequences. The Netherlands’Government has chosen to wait for results fromtest sites before committing to significant develop-ment of offshore windfarms.

ESC10 9/5/06 3:09 PM Page 377

378 PIERS LARCOMBE

10.4 SHELF SEDIMENTARY SYSTEMS AND MANAGEMENT

10.4.1 National and international measures

There is a wide range of international and industrymeasures aimed at protecting the marine envir-onment. Some of the international agreements,conventions and legislation apply to offshoreactivities, some of which have relevance to shelf sedimentary systems. Especially notable is OSPAR, the Convention for the Protection of the Marine Environment of the North EastAtlantic (www.ospar.org). OSPAR requires theapplication of the precautionary principle, thepolluter-pays principle, best-available techniquesand best environmental practice, including the useof clean technology. Three annexes to the Con-vention relate to the prevention and eliminationof pollution from land-based sources, by dump-ing or incineration and from offshore sources,and two others relate to assessing the quality ofthe marine environment, and protection and con-servation of ecosystems and biological diversity.

Other conventions are also relevant in a sedi-mentary context, such as the United NationsConvention on Biodiversity. As part of imple-menting this convention, countries are required,for example, to prepare Biodiversity Action Plans,which involves some mapping of their continentalshelves (see e.g. http://www.jncc.gov.uk).

10.4.2 Types of impacts on shelf habitats

Shelf sedimentary systems form the physicalstructure for shelf habitats. This structure canbe impacted by:1 Physical loss, for example by removal oractivities that may influence the sedimentaryregime and hence change the sea bed.2 Physical damage, for example by dredging,bottom trawling and extraction. Some habitatsmay be more resilient and recover faster thanothers, but extensive physical damage may leadto loss of the original habitat.3 Non-toxic contamination, such as by enrich-ment by nutrients or organic matter. Physical

Fig. 10.13 Swath-bathymetric image of a nearshore sand bank in the North Sea. The dark oval depressions mark the location ofindividual monopiles, which are 4.2 m in diameter and 350 m apart, around which have developed scours up to 5 m deep and 50 m indiameter. There are no scours or raised areas are continuous between adjacent monopiles and there is continuity of the crests of largesubaqueous dunes across the sandbank and near the monopiles. Thus in this case, the primary physical evidence indicates that themonopiles appear to act individually on the sea-bed rather than in combination. (Image courtesy of EoN (Coventry), Andrews (Great Yarmouth) and CEFAS.)

ESC10 9/5/06 3:09 PM Page 378


disturbance by fishing may cause changes innutrient cycling in some marine ecosystems(Duplisea et al. 2001), which may be importantin avoiding eutrophication or anoxia in or nearthe sea-bed.4 Biological disturbance, e.g. introduction ofnon-native species, although of the 53 non-native species recorded in British waters, mostare causing concern in estuaries and coastalenvironments, rather than on the open shelf.5 Climate change, whereby long-term changes inenvironmental conditions (temperature, salinity,etc.) may result in changes in the distributionand nature of some marine species assemblages.Although not necessarily negative, a changedsea-bed flora or fauna may lead to changes in sea-bed stability or biogeochemical cycling. Wherebiogenic sediments are significant contributorsto present sediment budgets, or the degree andnature of bioturbation are changed, there is thepotential for some change to sedimentary condi-tions at the sea bed.

In terms of inputs of sedimentary material tothe shelf system, issues fall into three categories:1 volume – i.e. the presence of material thatmay cause a problem to navigation or whichmay tend to cause bathymetric changes deemedundesirable;2 composition – especially contamination, wherethe sediment chemistry or biology may harm the receiving environment (including sedimentorganic content);3 texture – which may be sufficiently differentfrom the naturally occurring material to bedeemed undesirable.

10.4.3 Environmental remediation

A working definition of remediation is ‘theaction taken at a site subjected to anthropogenicdisturbance to restore or enhance its ecologicalvalue’, of which the main categories are:1 Physical removal of material (dredging), which,as noted elsewhere, is mostly an issue in coastaland estuarine environments.2 Physical isolation of the material (contain-ment), which may include capping of material in situ (Palermo et al. 1998). Capping is the

physical covering of disposed material with other(cleaner or coarser) material in order to limit itsrelease and its exchange with the wider marineenvironment. There is an ongoing debate aboutthe long-term effectiveness of capping, particu-larly in deep water and in high-energy environ-ments. Capping can be viewed as changing theproblem rather than solving it, but can fulfilshort- and medium-term needs.3 Chemical treatment of the sediment.

Amongst others, remediation covers the fol-lowing management options.1 Non-intervention, whereby natural processesare allowed to proceed without human interven-tion (as opposed to, for example, the exclusionof activities such as trawling from occurringnearby).2 ‘Restoration’, whereby the ecosystem is re-turned to the condition or state that would exist had no dredging occurred. In practice, thisis rarely achievable. Athough it might be con-sidered relatively easy to define a general goalsuch as ‘restoration’, the complexity and dynamicnature of marine systems mean that there are genuine difficulties in establishing specific objectives that will be scientifically credible andmeasurable.3 Rehabilitation, where some of the original eco-logical features are replaced by different ones.4 Habitat enhancement or creation, where theoriginal ecosystem is replaced by another, eitherat the site of impact or elsewhere.

The need for remediation may depend uponGovernment policy, or be assessed on a case-by-case basis. Clearly, the sedimentary dynamics ofthe receiving environment must be well under-stood, so allowing an assessment to be made ofthe technology used for material emplacementand the likely and acceptable levels of naturaldisturbance and dispersal. In terms of sedimentdynamics, receiving environments tend to fallbetween two end members:1 low-energy environments, quiescent and withrelative long-term stability;2 high-energy environments, dynamic and poten-tially dispersive.It is not always clear whether natural retentionor dispersal is the best environmental outcome.

ESC10 9/5/06 3:09 PM Page 379

380 PIERS LARCOMBE

10.4.4 ‘Beneficial use’ of dredged material

Dredged material can be a valuable resource,but at present, in the UK, only c. 1% of materialdredged for disposal at sea is currently reused inthe marine environment. This may be higher inother countries but data are not easily available.In many countries, research is being undertakento identify ways of increasing the percentage ofdredged material used in beneficial use schemes.Use of such material on the shelf is minimal, and includes shoreline protection by depositingmaterial on the inner shelf (Small et al. 1997).Most other beneficial use schemes are associatedwith the coastline, including habitat restoration,the maintenance of coastal sediment cells and the(re-)creation of beach or saltmarsh habitats (e.g.the DECODE project, http://www.cefas.co.uk/decode/use.htm). The joint website of the U.S.EPA and U.S. Army Corps of Engineers listsdetails of a number of valuable case studies(http://el.erdc.usace.army.mil/dots/budm/).

10.4.5 Managing impacts

Around the world there are generally four mainimpacts to manage. Regarding material inputsto continental shelves, the disposal of dredgespoil is the main issue, and regarding sea-beddisturbance, the main issues relate to trawling,aggregate dredging and marine mining. Dredgespoil and aggregate dredging are consideredbelow.

10.4.5.1 Dredge spoil disposal

In the USA, the U.S. Army Corps of Engineers(USACE) and the U.S. Environmental Protec-tion Agency (EPA) have statutory responsibilitiesfor the management of dredged material place-ment in ocean, inland and nearshore waters.When considering open-water placement ofdredged material, the potential for water col-umn and benthic effects related to sediment contamination is to be evaluated, and manage-ment options considered that aim to reduce therelease of contaminants to the water column

during placement and/or subsequent isolationof the material from benthic organisms. Suchoptions include operational modifications, use ofsubaqueous discharge points, diffusers, subaque-ous lateral confinement of material, or cappingof contaminated material (Francingues et al. 1985; USACE/EPA 1992). Currently, the broadresearch programme Dredging Operations andEnvironmental Research (DOER 2005) is under-way, conducting research designed to providedredging project managers with technology toimprove cost-effective operation, evaluation ofrisks associated with management alternativesand environmental compliance.

Canada has a range of measures in place to pro-tect its marine environment, under the CanadianEnvironmental Protection Act, 1999 (CEPA), andto meet international commitments. EnvironmentCanada conducts representative monitoring atsites of disposal at sea, and in 2003, monitor-ing activities were conducted at 17 disposal sites (London Convention 2005; www.ec.gc.ca/seadisposal/) involving assessments of the phys-ical, chemical and biological characteristics. Per-mits for disposal include the generation of ‘impacthypotheses’, which form the basis of subsequentmonitoring. Physical monitoring involves thecollection of relevant geological information fordetermining the area of deposition, delineatingthe disposal-site boundaries, studying the accu-mulation of dredged material within the area ofdeposition and documenting evidence of sedi-ment transport from the disposal site. Associatedbiological and chemical assessments are under-taken, the monitoring design for which takes intoaccount the site’s size and dispersal characteris-tics. Major sites of disposal (> 100,000 m3 yr−1

of dredged material) are monitored at least every5 years, and monitoring of other sites is based onvolume, proximity to sensitive areas, or level ofpublic concern. In common with many monitor-ing programmes, Canada’s programme dependson funds gained from the collection of fees frompermittees. In the UK, annual Aquatic Envir-onment Monitoring Reports (AEMR) include sections on marine monitoring activities (availablefrom www.cefas.co.uk).

ESC10 9/5/06 3:09 PM Page 380


10.4.5.2 Aggregate dredging sites

Risk assessment and management are basedbroadly on maximizing the potential for post-cessation change back to pre-existing conditions.The basic techniques include:1 minimizing the area of impact and dredgingintensities, which determines the starting pointfor ‘recovery’;2 maximizing time-gaps between successivedredging events (c. 6 years can produce near re-establishment, but it depends upon the sediment-ary dynamics), gaps between individual dredgingfurrows and the depth of individual furrows;3 using a site-specific assessment of impact and recovery, based on ‘types’ of gravel bioticassemblages;4 leaving a thickness of at least 0.5 m of com-parable coarse substrates (Boyd et al. 2004).

For UK waters, a valuable management toolis the Electronic Monitoring System (EMS),which automatically records the geographicallocation and duration of dredging and dis-posal activity on board each dredging vessel(Fig. 10.14). Such data form a valuable set that

can be used to inform risk assessments of planned,active and ceased dredging activities. For onebank off southern England, EMS data showedthat dredging intensities in some areas reached upto 5–15 hours per year per 100 m2 box (Boyd et al. 2004). Note that fishing trawlers over acertain size on the European UK continentalshelf are also subject to EMS.

A range of remediation activities related toaggregate dredging are undertaken in UK waters,including minimizing the total area licensed/permitted for dredging, and carefully locating anynew dredging areas with regard to the findingsof an Environmental Impact Assessment. Further,dredging practices are adopted that minimizeimpact and operators are required to monitor theenvironmental impacts of their activities. Finally,dredging operations are controlled through theuse of conditions attached to the dredging licence.As a result, basic strategies for remediation foraggregate dredging are developing rapidly (e.g.EMU 2004). Guiding principles include drawingupon the best scientific advice possible to estab-lish (i) the need for remediation, (ii) the generalgoals and specific objectives of remediation and,

0 (km) 100

Area 22

South-eastEngland

Dredging intensity in hours in each 100 m x 100 m block

<1.00

1.00–1.99

2.00–2.99

3.00–3.99

4.00–4.99

>5.00

200 m

Fig. 10.14 Dredging intensity in hours per 10,000 m2 block for a site off Felixstowe (Area 422), south-east England for the year 1995.Dotted line indicates boundaries of disposal ground. (Adapted from Boyd et al. 2003.)

ESC10 9/5/06 3:09 PM Page 381

382 PIERS LARCOMBE

importantly, (iii) the criteria by which perform-ance will be measured.

10.4.6 Contaminants and sediment quality guidelines

As part of an assessment of environmental impact,many countries have developed Sediment QualityGuidelines, which are used to inform decisionstaken about sediment management (Chapter 1;Wenning et al. 2005). A number of legislativedrivers require consideration of the potentialimpact of contaminated sediment on aquaticenvironments, including, in Europe, the HabitatsDirective and the Water Framework Directive.The guidelines are used to assess the quality ofmaterial that may be dredged (mostly from harbours and coastal waters) and the quality of sediments (e.g. dredge spoil) placed into the shelfenvironment. In assessing dredged material andits suitability for disposal to sea, defined concen-trations of contaminants (‘Action Levels’) arenot generally used as a simple pass/fail test, butrather are used as part of a ‘weight of evidence’approach (CEFAS 2003). This reflects recentweight-of-evidence approaches to environmen-tal management of sediments, whereby multiplelines of evidence concerning ecological assessmentare used as an aid to decision making (Chapman1986, 1996; Burton 2002). One relatively well-developed approach to setting sediment qualitycriteria is that of Long et al. (1995) and Long & MacDonald (1998) in the USA. A statisticalapproach matches biological and chemical datafrom laboratory and field studies in NorthAmerica, and from modelling work, and the tech-nique has since been developed to set sedimentquality guidelines in a number of countries,notably the USA, Canada and Hong Kong.

In the UK, applications for the dredging anddisposal of contaminated material must con-sider alternative options, including placing the material in landfill, ‘capping’ it in the marineenvironment and forming a new reclamationsite beneath which the material is placed (i.e. an engineered new site). Much related researchhas been performed by the U.S. Army Corp ofEngineers (DOER 2005). On occasion, high costscan lead to a further option, of doing nothing. In

forming action levels, the philosophical approachtaken is to use sediment chemistry to provide an estimate of background metal concentra-tions for Action Level 1 (Table 10.4). The deriva-tion of Action Level 2 concentrations has beenguided by ecotoxicological information from theliterature, applied to concentrations in samplessent in support of applications. Limited eco-toxicological data are available, and most suchwork has been done in North America (Burton 2002). Work is ongoing to derive background‘reference concentrations’ for England and Wales,so the concentrations given here are likely to be revised. In Europe, under OSPAR, there is no consistent approach taken; for example,countries analyse different size fractions, somecountries may digest their samples only partiallyand others digest samples fully using hydrogenfluoride. Although this largely relates to coastaland estuarine sites, it is indicative of some of theproblems involved, which could be applicable to shelf environments.

There are also concentrations that are used to assess sediment recovered in post-disposalmonitoring programmes, including those formetals (e.g. Rowlatt et al. 2002) and for organiccompounds. Environment Canada has a seriesof lower ‘Action Levels’ (London Convention2005) for sediments collected during monitoringprogrammes, which form part of an integratedassessment procedure (Table 10.5). If sedimentsare below the lower action levels for contami-nants and pass all biological tests, no furtheraction is required. If levels of contaminants orbiological test results demonstrate a cause forconcern, however, then:1 compliance is verified with the terms of thepermits issued since the site was last monitored;2 potential sources of pollutants are checkedand further site characterization undertaken.Cursory benthic community surveys can be used as a general sediment quality indicator, but the overall assessment of the disposal site con-siders all information available from physical,chemical and biological monitoring.

Within Europe, international co-operation onsuch issues is not finalized, but the indication is that monitoring of disposal sites should aim

ESC10 9/5/06 3:09 PM Page 382


to determine temporal trends rather than testfor compliance, and that the aim is for no deter-ioration in sediment quality. There are otherapproaches being assessed in some countries,such as the ‘Added Risk’ approach, where back-ground concentrations are determined and atoxicological quotient is added to derive a maximum permissible concentration. For thosecontaminated sediments deemed unsuitable forconventional disposal, they may be confined, con-tained, treated or simply not dredged.

10.4.7 Design of research and monitoringprogrammes

There is a very wide range of advice documentsavailable (mostly on the web) regarding varioushuman activities on the shelf that have sedi-mentary impacts, and there are a suite of generalthemes that should be considered in designingresearch and monitoring for shelf sedimentarysystems. It is clearly important to monitor at frequencies and with accuracies that will reveal

information on the relevant shelf sedimentaryprocesses. Given the complexities of monitoringsediment transport processes, this can be expen-sive. Increasingly, real-time or near real-time dataare broadly available on the main oceanographicparameters, through outputs from regional mon-itoring programmes (e.g. in the USA, Europeanwaters and the Great Barrier Reef shelf). It isalso sensible to obtain data across spatial scalessufficiently broad to assess the driving forces

Table 10.5 Canadian Environmental Protection Agency(CEPA) Lower Action Levels (from Disposal at Sea Regulations)for selected chemicals in sediments (in mg kg−1 dry weight)(London Convention 2005).

Chemical Current level

Cadmium 0.6Mercury 0.75Total PCBs 0.1Total PAHs 2.5

PCB, polychlorinated biphenyls; PAH, polycyclic aromatichydrocarbon.

Table 10.4 Action levels (mg kg−1 dry weight) for metals in sediments in the USA, Canada, Hong Kong, ICES Working Group onMarine Sediments (ICES WGMS) and for England and Wales. (ICES, International Council for Exploration of the Sea; ERL, Effects RangeLow; TEL, Threshold Effects Level; ERM, Effect Range Median; PEL, Probable Effect Level; ISQV, Interim Sediment Quality Value). (From CEFAS 2003.)

Metal Ecotoxicological standards Chemistry (mainly)

NOAA Canada Hong Kong ICES WGMS England and Wales existing guidelines guidelines range of values (approximate value)

ERL TEL ISQV-low Action Level 1As 8.2 7.24 8.2 20–80 20Cd 1.2 0.676 1.5 0.5–2.5 0.4Cr 81 52.3 80 60–300 40Cu 34 18.7 65 20–150 40Hg 0.15 0.13 0.28 0.1–1 0.3Ni 20.9 15.9 40 37–130 20Pb 46.7 30.2 75 30–120 50Zn 150 124 200 160–700 130

ERM PEL ISQV-high Action Level 2As 70 41.6 70 50–1000 50–100Cd 9.6 4.21 9.6 2.4–12.5 5Cr 370 160 370 180–5000 400Cu 270 108 270 90–1500 400Hg 0.71 0.7 1 0.8–5 3Ni 51.6 42.8 – 45–1500 200Pb 218 112 218 100–1500 500Zn 410 271 410 500–10000 800

ESC10 9/5/06 3:09 PM Page 383

384 PIERS LARCOMBE

of the relevant sedimentary processes. Maps of the sea-bed facies are essential backgroundinformation and also provide the basis for theclassification of sea-bed habitats. Clearly, it isdesirable that programmes should be availablefor scientific scrutiny, so that the outcomes aretransparent and scientifically defensible. It isincreasingly important that information and dataare rapidly disseminated. Finally, the importanceof performance criteria cannot be overempha-sized – there may be little point in undertakingwork unless progress towards the objectives canbe measured and the effectiveness of manage-ment actions can be assessed.

There are good recent examples of shelf map-ping projects. Extensive mapping work is beingundertaken in Australia (www.oceans.gov.au/auscan), which was designed to assist theecosystem-based approach to environmentalmanagement (section 10.4.8) by improvingunderstanding of the links between the physicaland biological components. Work in the IrishSea is developing a physical basis for the con-servation of marine habitats (The Joint NatureConservation Committee 2004). The Irish Na-tional Sea-bed Survey is also underway; usingtechniques including multibeam and single-beamecho sounders, sub-bottom profiling, gravity,magnetics, with ground-truthing performed usingbox core samples (http://www.gsisea-bed.ie). A pan European project (Mapping EuropeanSea-bed Habitats (MESH) aims to produce sea-bed habitat maps for much of the north-west European shelf, and will improve the inputof such data into environmental managementwithin national regulatory frameworks (http://www.searchmesh.net). Complementary regionalstudies of the composition of shelf sediments(e.g. Stevenson 2001) are also useful, especiallywhen dealing with issues of disposal at sea.

10.4.8 Taking an ecosystem-based approach

Marine systems are naturally dynamic and dis-play natural variability, and the human pressureson them vary with different patterns of humanactivity. Protection of marine environmentsneeds to be flexible and adaptable. Although

by no means universal, there is increasingly an‘ecosystem’ approach taken to environmentalmanagement of many ecosystems, includingthose on continental shelves (e.g. EuropeanMarine Strategy 2005). Essentially, such anapproach means that, rather than viewing ahuman impact and its management in isolation,account is taken of the range of components ofthe system in question, the relevant processesand interactions between components, the envir-onmental factors that influence the system andthose factors that might be affected by anthro-pogenic intervention. For example, except foraggregates or marine minerals, the distributionof shelf sediments and their associated processesis not generally seen as having an inherent value. Rather, sediments are important as a substrate for benthic organisms and as a com-ponent of the broader marine ecosystem. Thus,assessment and monitoring of sediments areoften performed with regard to their role informing and maintaining marine habitats.

The ecosystem-based approach requires anunderstanding of the overall sedimentary phys-ical system. Most continental shelves are hugephysical systems, so that there is little practicaldifference most human activities can make to thesystems as a whole, at least in terms of sedimentbudgets. Consequently, local and subregionalscales tend to be most relevant, with biological,chemical and social factors of importance. Map-ping the sea-bed is a high priority for the future.

10.5 FUTURE ISSUES

10.5.1 Challenges of managing the impacts ofhuman activities on shelf sedimentation

Study, assessment and management of theimpacts of human activities on the sediments of continental shelf environments represent asignificant challenge. The reasons include themany physical processes that influence the shelfsedimentary environment (Figs 10.2 & 10.3),ranging from turbulence (scales of seconds andcentimetres) through to major sea-level cycles(scales of c. 100,000 yr and c. 100 m). These

ESC10 9/5/06 3:09 PM Page 384


directly influence chemical and biological pro-cesses and operate on a variety of spatial andtemporal scales, which may be hard to sampleeffectively. In addition, where two or more humanactivities coincide to influence the sediments, thecombined impact may be a complex function ofthe nature of the individual impacts (e.g. trawl-ing in an area of aggregate dredging). There are also practical difficulties and great expenseinvolved in obtaining information about theshelf sea-bed and its sedimentary processes.

Further, management regimes need to bebased on the best available science, so scienceknowledge needs to be transformed effectivelyinto policy and management actions to preventenvironmental damage. Very real constraintsexist on our ability to do this, not least becauseof the difficulty of proving cause and effect, andof dealing with the uncertainties involved. It isalso noteworthy that science knowledge is onlyone of a broad range of factors that can be takeninto account when forming policy (Fig. 10.15)

10.5.2 Sedimentary impacts of sea-beddisturbance

The main current issues of anthropogenic sea-bed disturbance were examined in above sec-

tions, some of which are likely to be highly rele-vant in the future.

10.5.2.1 Trawling

Many countries’ economies and populations relyheavily upon trawling on their continental shelves,for example, in 2001, the Great Barrier Reef(GBR) shelf was worth AU$4.2 million to theAustralian economy, and another AU$0.3 mil-lion through recreational and commercial fish-ing (Productivity Commission 2003). On someshelves, trawling is already heavily regulated,primarily to protect fish-stocks or allow theirrecovery, but with the added advantage of moderating impacts upon the sea bed. Movestowards enforcing marine zonation of varioustypes are also occurring. Perhaps the mostadvanced in this process is the GBR shelf, wherethe wish to minimize potential human impactshas led to the shelf being divided up into a seriesof complex biophysical zones, within whichrestrictions apply to various human activitiessuch as anchoring, line fishing and trawling.Such zonation is driven by perceived ecosystemconsequences rather than sedimentary ones perse, and there remains a need for fundamentalresearch relating the shelf sedimentary environ-ment to the ecology.

10.5.2.2 Aggregate dredging

World-wide, the demand for marine aggregatesis largely a function of construction activity. For the UK, most marine sand and gravel is currently extracted from the east and south-east coasts of England, but new discoveries inthe eastern English Channel, near the medianline with France, are planned for exploitation.The UK’s past and predicted requirement isaround 25 × 106 t yr−1 until 2016, of which thesenew deposits could provide > 50% (DEFRA2005). Licence conditions include commit-ments towards environmental protection andmonitoring, so that trends in the sedimentaryimpacts of marine aggregate extraction will besupported by increasing volumes of relevantsedimentary data.

Policy

Ideas

Joined-upworking

TimingEvidence

Practicality/delivery

Devolution

Science andtechnology

LegalityCost andresources

Global pressuresEUobligations

Dogma

Organizational culture

Ministers'ambition/

ability/willpower

Specialistlobbying

The media

Think tanks

Public opinion

Parliament

Events and crises

Public commitmentsand targets

The Manifestoand party pressure

Fig. 10.15 Graphic illustrating 22 factors that can influence the decisions taken by UK Government ministers. Note that‘science and technology’ forms only one factor. (Used withpermission of McNeil Robertson Development.)

ESC10 9/5/06 3:09 PM Page 385

386 PIERS LARCOMBE

A Regional Environmental Assessment (REA)of the eastern English Channel has focused upon those sea areas that may be impacted bydredging. The REA includes predictions ofdredging-related physical and biotic changes,especially related to recolonization of the bed,but the evidence base for such predictions requiresdevelopment. The findings of field-based studiesdiffer, ranging from minimal effects of distur-bance following cessation of dredging (Robinsonet al. 2005) to significant changes in communitystructure which persist over many years (Desprez2000). Although some of these differences prob-ably reflect the different sediment types and theirassociated different physical dynamics, compo-sition and biogeochemical cycling, links betweenpost-dredging ecological change and physicalprocesses need to be strengthened in order toallow better transfer of conclusions to other areas.

10.5.2.3 Offshore renewable energy developments(OREDs)

For countries with continental shelves, the renew-able energy resource can include tides, currents,waves and offshore wind. The development ofoffshore windfarms is now well underway insome countries (section 10.3.2.5.). In December2003, the UK Government announced a com-mitment to generating 10% of the UK electricityneeds from renewable sources by 2010 and20% by 2020. Most of this is likely to be met by a major increase in the number and size ofoffshore windfarms. By late 2005, there were 15 projects awarded and in the early stages ofplanning, which will amount to a generationcapacity of 7.2 GW, potentially contributingelectricity to more than 4 million households.Other types of development are increasinglyunder consideration, including underwater tidal turbines and floating offshore windfarms(Pelc & Fujita 2002; Gill 2005). With offshorestructures, the likely impacts upon the shelf sedimentary system are generally local in nature.Some offshore structures, especially those cover-ing extensive areas such as windfarms, causelocal disturbance, especially from scour aroundfoundations and the associated impacts upon

local habitats (Rees, in Judd et al. 2003), yetmay also be viewed as having positive effects iftheir presence reduces or halts trawling in thearea (Elliott & Cutts 2004).

Extraction of wave energy is also a rapidlyexpanding field, and the UK is one of the richestnations in terms of potential for wave energy(Pelc & Fujita 2002). Extraction of wave energyreduces the energy reaching the coast, andwhether this effect is perceived as beneficial ordetrimental depends on the specific coastline.There are no significant impacts likely on shelfsediments, however, unless there is an import-ant active link between the coastal and shelf sediments. Tidal flows are also a huge potentialenergy source. World-wide, tidal power plantscould theoretically generate 500–1000 terraWatthours per year (TWh yr−1, terra = 1012) andwestern Europe 100 TWh yr−1. Theoretically theUK could generate 50 TWh yr−1, 125 times thecurrent UK use (0.4 TWh yr−1, www.dti.gov.uk)although economic constraints mean that only afraction of this energy is likely to be exploitable(Pelc & Fujita 2002). Tidal turbines are arraysof impellors located above the bed, generally in locations where tidal current speeds exceed c. 2 m s−1, and appear to have few anticipatedsedimentary impacts.

10.5.3 Sedimentary impacts of climate change

For continental shelves, it is probably true thatbasic knowledge of natural sedimentary changeson time-scales of decades is mostly insufficientto allow anthropogenic changes to be confid-ently identified and their significance assessed.Studies of marine sediments and fossils can helpaddress whether recently measured shelf sedi-mentary changes have occurred before. Althoughthe details of cause and effect may often beimperfectly understood, better links are possiblebetween the understanding of the natural driv-ing forces and the tools used for management(Hardman-Mountford et al. 2005; Larcombe et al. 2005; Rogers & Greenaway 2005). Withregard to shelf sediments, the main impacts ofclimate change include results of changes in sea-level and ocean circulation.

ESC10 9/5/06 3:09 PM Page 386


10.5.3.1 Impacts of sea-level change

Over the next few decades, changes in mean sea-level (of up to a few decimetres) may affectsome parts of the world’s continental shelves.For example, the vertical extent of many coralreefs is presently limited by sea-level, the reefshaving caught up with sea-level during the late Holocene (Larcombe & Woolfe 1999). As a result, a rise in relative sea-level may increasecoral cover on reef flats and thus increase ratesof carbonate production. In deeper shelf areas,changed sea-level will have little impact oversuch time-scales, because here marine processesdominate sedimentation, and changed oceaniccirculation patterns, with consequent changes in temperature and storminess, may be moreimportant.

10.5.3.2 Impacts of ocean current change

Changes in shelf sedimentation directly relatedto varying ocean currents have been notedabove (section 10.2.3.5 and Case Study 10.2).Ocean currents may influence shelf sedimenta-tion in other ways, through their introduction of water of varying temperature and salinity,and the consequent potential for altered shelfsediment production. That changes in biogenicsediment production can occur as a result ofoceanographic changes is established on time-scales of centuries and millennia (e.g. Scourse et al. 2002), but equivalent data are lacking onshorter time-scales in shelf environments. Onsome shelves, there are recognized changes inshelf ecology related to variations in inputs ofoceanic waters (e.g. Beaugrand 2004), whichmight be expected to change primary produc-tion and the supply of organic matter to the sea bed.

10.5.3.3 Impacts of weather changes (floods, winds,storms)

Shelf sedimentation is likely to be changed if the relative magnitudes between parameters ofsupply, translation and removal are altered.Thus, if major changes in flood frequency occur,

there is the potential for significantly changedsediment supply. Reservoir construction is prob-ably the most important influence on sedimentsupply, but in many datasets it is difficult to dis-tinguish the influence of climate change fromthat of other changes in catchment conditions(Walling & Fang 2003). Where shelf sedimentdistribution depends upon waves and/or wind-driven currents (e.g. the GBR shelf), changedseasonal wind patterns have the potential tochange the distribution of shelf sediments. Forstorm-dominated shelves, changed frequency,magnitude and preferred storm tracks mightinfluence the distribution of shelf sediments.Overall, such oceanographic shifts and associatedecological changes are increasingly acknowledged,but the consequences on the sedimentology arenot understood.

10.5.4 Research needs and gaps

For continental shelves in general, and per-haps especially for temperate shelves, there is a lack of detailed information on past shelf‘weather and climates’, i.e. the short-, medium-and long-term variation in shelf processes.Knowledge is lacking on the dynamics of coarsesediments, and the role of such sediments inshelf processes (e.g. C and N cycling). Signific-ant limitations in understanding exist regard-ing the rates of various processes (see Becker et al. 2001), including sediment transport itselfand related biogeochemical processes. It is stilldifficult to make real-time observations of thenature of suspended particulates (Sternberg & Newell 1999). Regarding contamination,research is needed on so-called ‘combinedeffects’, whereby the detrimental effects of acontaminant are altered if other chemicals orenvironmental conditions occur. This is a verydifficult issue because of the rapid increase in the number of new chemicals.

Although geophysical survey devices (e.g.swath-mapping) are improving our knowledgeof the spatial distribution of various sediment-ary facies and biotopes, there remains a lack ofbasic understanding of the links between shelfsedimentary processes and benthic organisms,

ESC10 9/5/06 3:09 PM Page 387

388 PIERS LARCOMBE

especially for sandy and gravelly sediments. Field-based work is needed to document these links.Overall, we can anticipate increasing anthro-pogenic influences upon some parts of shelf sedimentary systems, although remaining over-all at a low level for most shelves. Case studiesare required that describe the sedimentary con-sequences of human intervention over sufficientlylong time-scales to allow the design of reliableand scientifically credible indicators of change.For some areas and/or types of human activity,environmental monitoring data of good qualitywill be available, including time-series data.Two fundamental and linked questions remain:

How do we distinguish anthropogenic impactsfrom natural variation, and assuming that wecan do this, how do we then ascribe a significanceto any differences? These questions are the focus of much ongoing and future research, inthe scientific, management and socio-economicspheres.

ACKNOWLEDGEMENT

This work was funded by the UK Departmentfor Environment, Food and Rural Affairs con-tract numbers A1225 and E3203.

ESC10 9/5/06 3:09 PM Page 388

Aalto, R., Maurice-Bourgoin, L., Dunne, T., et al. (2003)Episodic sediment accumulation on Amazonian floodplains influenced by El Niño/Southern Oscillation.Nature 425, 493–7.

Acker, K.L. & Stearn, C.W. (1990) Carbonate–siliciclasticfacies transition and reef growth on the northeastcoast of Barbados, West Indies. Journal of SedimentaryPetrology 60, 18–25.

Ackers, P. & Thompson, G. (1987) Reservoir sedimenta-tion and the influence of flushing. In: Sediment Trans-port in Gravel Bed Rivers (Eds C.R. Thornes, J.C.Bathhurst & R.D. Hey), pp. 845–68. Wiley, Chichester.

Adam, P. (1990) Saltmarsh Ecology. Cambridge Uni-versity Press, Cambridge, 461 pp.

Alam, I.A. (1993) The 1991 Gulf war oil spill: lessonsfrom the past and a warning for the future. MarinePollution Bulletin 27, 357–60.

Alexander, D. (1998) Natural Disasters. UCL Press,London, 632 pp.

Allen, J.R.L. (1964) Studies in fluviatile sedimentation:six cyclothems from the Lower Old Red Sandstone,Anglo-Welsh Basin. Sedimentology 3, 163–98.

Allen, J.R.L. (1970) A quantitative model of grain sizeand sedimentary structures in lateral deposits. Geo-logical Journal 7, 129–46.

Allen, J.R.L. (1970) Physical Processes of Sedimentation.Allen and Unwin, London, 248 pp.

Allen, J.R.L. (1984) Sedimentary Structures, their Char-acter and Physical Basis. Elsevier, Amsterdam.

Allen, J.R.L. (1985) Physical Processes of Sedimentation.Allen and Unwin, London.

Allen, J.R.L. (1997) Simulation models of salt marsh mor-phodynamics: some implications for high intertidalsediment couplets related to sea level change. Sedi-mentary Geology 113, 211–23.

Allen, J.R.L. (2000) Morphodynamics of Holocene salt marshes: a review sketch from the Atlantic andsouthern North Sea coasts of Europe. QuaternaryScience Reviews 19, 1155–231.

Allen, P. (1997) Earth Surface Processes. BlackwellScience, Oxford, 404 pp.

Aller, R.A. & Blair, N.E. (1996) Sulfur diagenesis andburial on the Amazon shelf: major control by physicalsedimentation processes. Geomarine Letters 16, 3–10.

References

Aller, R.C., Hannides, A., Heilbrun, C., et al. (2004)Coupling of early diagenetic processes and sediment-ary dynamics in tropical shelf environments: the Gulfof Papua deltaic complex. Continental Shelf Research24, 2455–86.

Allison, M.A., Nittrouer, C.A., Faria, L.E.C. Jr., et al.(1996) Sources and sinks of sediment to the Amazonmargin; the Amapa coast. Geomarine Letter 16, 36–40.

Allison, M.A., Khan, S.R., Goodbred, S.L.Jr., et al.(2003) Stratigraphic evolution of the late HoloceneGanges-Brahmaputra lower delta plain. SedimentaryGeology 155, 317–42.

Allott, R.W., Hewitt, C.N. & Kelly, M.R. (1990) Theenvironmental half-lives and mean residence times of contaminants in dust for an urban environment:Barrow-in-Furness. The Science of the Total Environ-ment 93, 403–10.

Alsharhan, A.S. & Kendall, C.G.st.C. (2003) Holocenecoastal carbonates and evaporites of the southernArabian Gulf and their ancient analogues. Earth-Science Reviews 61, 191–243.

Ambers, R.K.R. (2001) Using the sediment record in a western Oregon flood-control reservoir to assess theinfluence of storm history and logging on sedimentyield. Journal of Hydrology 244, 181–200.

Ambio (1976) Special issue on acid rain. Ambio 5(5–6).Ambio (1990) Special issue on marine eutrophication.

Ambio 19, 102–76.Amos, K.J., Alexander, J., Horn, A., et al. (2004) Supply

limited sediment transport in a high-discharge event of the tropical Burdekin River, North Queensland,Australia. Sedimentology 51, 145–62.

Andrews, J.E., Pedley, H.M. & Dennis, P.F. (1994)Stable isotope record of palaeoclimatic change inBritish Holocene tufa. The Holocene 4, 349–55.

Andrews, J.E., Brimblecombe, P., Jickles, T.D., et al.(1996) An Introduction to Environmental Change.Blackwell, Oxford, 209 pp.

Ankley, G.T., Di Toro, D.M., Hansen, D.J., et al. (1996)Technical basis and proposal for deriving sedimentquality criteria for metals. Environmental Toxicologyand Chemistry 15, 2056–66.

Anthony, E.J. & Blivi, A.B. (1999) Morphosediment-ary evolution of a delta-sourced, drift-aligned sand

ESD01 9/5/06 3:09 PM Page 389

390 REFERENCES

barrier-lagoon complex, western Bight of Benin. MarineGeology 158, 161–76.

Anthony, E.J. & Julian, M. (1999) Source-to-sink sedi-ment transfers, environmental engineering and hazardmitigation in the steep Var River catchment, FrenchRiviera, southeastern France. Geomorphology 31,337–54.

Anthony, E.J. & Orford, J.D. (2002) Between wave- andtide-dominated coasts: the middle ground revisited.Journal of Coastal Research, Special Issue 36, 8–15.

Antoine, P. (1993) The Somme valley terrace system(northern France); a model of river response toQuaternary climatic variations since 800,000 B.P. TerraNova 6, 453–64.

Apitz, S. & White, S. (2003) A conceptual frameworkfor river-basin-scale sediment management. Journal ofSoils and Sediments 3, 132–8.

Appleby, P.G. & Oldfleld, F. (1979) Letter on the historyof lead pollution in Lake Michigan. EnvironmentalScience and Technology 13, 478–80.

Arizona Department of Transportation (1975) SoilErosion and Dust Control on Arizona Highways, 4 Vols. Arizona Department of Transportation,Phoenix.

Arnaud-Fassetta, G. (2003) River channel changes in theRhome Delta (France) since the end of the Little IceAge: geomorphological adjustment to hydroclimaticchange and natural resource management. Catena 51,141–72.

Ashley, R.M., Fraser, A., Burrows, R., et al. (2000) Themanagement of sediment in combined sewers. UrbanWater 2, 263–75.

Aubanel, A., Marguet, N., Colombani, J.M., et al. (1999)Modifications of the shoreline in the Society Islands(French Polynesia). Ocean and Coastal Management42, 419–38.

Aulitzky, H. (1994) Hazard mapping and zoning inAustria: methods and legal implications. MountainResearch and Development 14(4), 307–13.

Axtmann, E.V. & Luoma, S.N. (1991) Large-scale distribution of metal contamination in fine-grained sediments of the Clark Fork River, Montana, USA.Applied Geochemistry 6, 75–88.

Bailey, R.G., Zoltan, S. & Wiken, E.B. (1985) Ecologicalregionalization in Canada and the United States.Geoforum 16, 265–75.

Baker, V.R. (1994) Geomorphological understanding offloods. Geomorphology 10, 139–56.

Baker, V.R., Kochel, R.C., Patton, P.C., et al. (1983)Paleohydrological analysis of Holocene flood slack-water sediments. In International Conference on FluvialSedimentology, Glasgow, 1981, pp. 229–39. BlackwellScientific, Oxford.

Ballantyne, C.K. (2002) Paraglacial geomorphology.Quaternary Science Reviews 21, 1935–2017.

Baltzer, F., Allison, M. & Fromard, F. (2004) Materialexchange between the continental shelf and mangrove-

fringed coasts with special reference to the Amazon-Guianas coast. Marine Geology 208, 115–26.

Barnhardt, W.A. & Kelley, J.T. (1995) The accumulationof carbonate sediments on the inner shelf of Maine: amodern consequence of glaciation and sea-level change.Journal of Sedimentary Research 65, 195–208.

Barrell, J. (1912) Criteria for recognition of ancient deltadeposits. Geological Society of America Bulletin 23,377–446.

Barrie, J.V. & Collins, M.B. (1989) Sediment transporton the shelf margin of the Grand Banks of Newfound-land. Atlantic Geology 25, 173–9.

Barry, R.G. (1992) Mountain Weather and Climate, 2ndedn. Routledge, London.

Barsch, D. & Caine, N. (1984) The nature of mountaingeomorphology. Mountain Research and Development4(4), 287–98.

Bates, C.C. (1953) Rational theory of delta formation.Bulletin of the American Association of PetroleumGeologists 37, 2119–62.

Bates, R.L. & Jackson, J.A. (1980) Glossary of Geology.American Geological Institute, 2nd edn. Falls Church,Virginia.

Battarbee, R.W. (2000) Palaeolimnological approachesto climate change, with special regard to the biologicalrecord. Quaternary Science Reviews 19, 107–24.

Baudo, R., Giesy, J.P. & Muntau, H. (Eds) (1990) Sedi-ments: Chemistry and Toxicity of In-Place Pollutants.Lewis Publishers, Chelsea, MI, 405 pp.

Beaugrand, G. (2004) The North Sea regime shift: evid-ence, causes, mechanisms and consequences. Progressin Oceanography 60, 245–62.

Beaumont, P. (1989) Drylands. Environmental Manage-ment and Development. Routledge, London.

Becker, G.A., Brockmann, U.H., Damm, P., et al. (2001)Instruments and observational methods. In: Berichteaus dem Zentrum für Meeres- und Klimaforschung –Reihe Z: Interdisziplinäre Zentrumsberichte. Nr. 3.The Changing North Sea: Knowledge, Speculationand New Challenges. Synthesis and New Conceptionof North Sea Research (SYCON) (Eds J. Sündermann,S. Beddig, I. Kröncke, G. Radach & K.-H. Schlünzen),pp. 212–38. Zentrum für Meeres- und Klimaforschungder Universität Hamburg (available at http://www.rrz.uni-hamburg.de/SYKON/).

Beckwith, P.R., Ellis, J.B. & Revitt, D.M. (1986) Heavymetal and magnetic relationships for urban sourcesediments. Physics of the Earth and Planetary Interiors42, 67–75.

Beets, D.J. & van der Spek, A.J.F. (2000) The Holoceneevolution of the barrier and back barrier basins ofBelgium and the Netherlands as a function of the lateWeichselian morphology, relative sea level rise andsediment supply. Geologie en Mijnbouw 79, 3–16.

Belderson, R.H., Johnson, M.A. & Kenyon, N.H. (1982)Bedforms. In: Offshore Tidal Sands (Ed. A.H. Stride),pp. 27–57. Chapman and Hall, London.

ESD01 9/5/06 3:09 PM Page 390

REFERENCES 391

Belknap, D.F., Kelley, J.T. & Gontz, A.M. (2002)Evolution of the glaciated shelf and coast of the Gulfof Maine, U.S.A. Journal of Coastal Research (SpecialIssue) 36, 37–55.

Bellucci, L.G., Frignani, M., Paolucci, D., et al. (2002)Distribution of heavy metals in sediments of the VeniceLagoon: the role of the industrial area. The Science ofthe Total Environment 295, 35–49.

Bellwood, D.R., Hughes, T.P., Folke, C., et al. (2004)Confronting the coral reef crisis. Nature 429, 827–33.

Belperio, A.P. (1979) The combined use of wash loadand the bed material load rating curves for the calcula-tion of total load: an example from the BurdekinRiver, Australia. Catena 6, 317–29.

Belperio, A.P. (1983) Late Quaternary terrigenous sedimentation in the Great Barrier Reef lagoon. In:Proceedings of the Great Barrier Reef Conference (EdsJ.T. Baker, R.M. Carter, P.W. Sammario & K.P. Stark),pp. 71–6. James Cook University, Townsville.

Belperio, A.P. (1988) Terrigenous and carbonate sedimentation in the Great Barrier Reef province. In:Carbonate–Clastic Transitions (Eds L.J. Doyle &H.H. Roberts), pp. 143–74. Developments in Sedi-mentology 42, Elsevier, Amsterdam.

Benedict, J.B. (1970) Downslope soil movement in a Colorado alpine region. Rates, processes and climatic significance. Arctic and Alpine Research2(3), 165–226.

Bengtsson, Å. (1991) Effects of pollutants on vertebraeand spinal columns in the fourhorn sculpin (Myoxo-cephalus quadricornis L.) in the Gulf of Bothnia.Thesis, Umeå Univeristy, Sweden.

Benito, G., Baker, V.R. & Gregory, K.J. (1998)Palaeohydrology and Environmental Change. Wiley,Chichester.

Benjamin, M.M. & Leckie, J.O. (1982) Multiple-siteadsorption of Cd, Cu, Zn and Pb on amorphous iron oxyhydroxide. Journal of Colloid and InterfaceScience 79, 209–21.

Benzoni, F., Bianchi, C.N. & Morri, C. (2003) Coralcommunities of the northwestern Gulf of Aden(Yemen): variation in framework building related toenvironmental factors and biotic conditions. CoralReefs 22, 475–84.

Berendsen, H.J.A. & Stouthamer, E. (2000) Weichselianand Holocene palaeogeography of the Rhine-meusedelta, the Netherlands. Palaeogeography, Palaeo-climatology, Palaeoclimatology 161, 311–35.

Berger, A.R. & Iams, W.J. (1996) Geoindicators:Assessing Rapid Changes in Earth Systems. BalkemaPress, Rotterdam, 466 pp.

Berner, R.A. (1980) Early Diagenesis. A TheoreticalApproach. Princeton Series on Geochemistry, PrincetonUniversity Press, New Jersey, 241 pp.

Bernier, P., Guidi, J.B. & Bottcher, M.E. (1996) Coastalprogradation and very early diagenesis of ultramaficsands as a result of rubble discharge from asbestos

excavations (northern Corsica, western Mediterranean).Marine Geology 144, 163–75.

Beveridge, J.P. & Culbertson, J.K. (1964) Hyper-concentrations of suspended sediment. Proceedings ofthe American Society of Civil Engineers, Journal of theHydraulics Division 90 (HY6), 117–28.

Bickerton, M.A. (1995) Macroinvertebrates and theDetermination of In-river Flow Need. Report to the National Rivers Authority, Anglian Region,Peterborough.

Biggins, P.D.E. & Harrison, R.M. (1986) Chemical speciation of lead compounds in street dusts. Environ-mental Science and Technology 14, 336–9.

Binding, C.E., Bowers, D.G. & Mitchelson-Jacob, E.G.(2003) An algorithm for the retrieval of suspendedsediment concentrations in the Irish Sea from SeaWiFSocean colour satellite imagery. International Journalof Remote Sensing 24, 3791–806.

Bird, E.C.F. & Fabbri, P. (1993) Geomorphological andhistorical changes on the Argentina delta, Linguriancoast, Italy. GeoJournal 29, 428–39.

Bird, E.C.F. (1985) Coastline Changes: a Global Review.Wiley, Chichester.

Bird, E.C.F. (1993) Submerging coasts: the Effects ofRising Sea Level on Coastal Environments. Wiley,Chichester.

Bird, E.C.F. (1996) Beach Management. Wiley, Chichester.Bird, E.C.F. (1998) The Coasts of Cornwall. Alexander

Press, Fowey.Bird, E.C.F. (2000) Coastal Geomorphology: an

Introduction. Wiley Chichester.Bird, J.F. (2000) The impact of mining wastes on the rivers

draining into Georges Bay, north-east Tasmania. In:River Management: the Australian Experience (Eds S. Brizga & B.L. Finlayson), pp. 151–72. Wiley,Chichester.

Birkeland, P.W., Shroba, R.R., Burns, S.F., et al. (2003)Integrating soils and geomorphology in mountains – an example from the Front Range of Colorado.Geomorphology 55, 329–44.

Björnsson, H. (2003) Subglacial lakes and jökulhlaups in Iceland. Global and Planetary Change 35(3–4),255–71.

Blackwell, M.S.A., Hogan, D.V. & Maltby, E. (2004)The short-term impact of managed realignment on soilenvironmental variables and hydrology. Estuarine,Coastal and Shelf Science 59(4), 687–701.

Blanchon, P., Jones, B. & Kalbfleisch, W. (1997)Anatomy of a fringing reef around Grand Cayman:storm rubble, not coral framework. Journal of Sedi-mentary Research 67, 1–16.

Bloesch, J. & Burns, N.M. (1980) A critical review ofsedimentation trap technique. Schweizerische Zeitschriftfur Hydrologie 42, 15–55.

Bloesch, J. & Uehlinger, U. (1986) Horizontal sedi-mentation differences in a eutrophic Swiss lake.Limnology Oceanography 31, 1094–109.

ESD01 9/5/06 3:09 PM Page 391

392 REFERENCES

Boorman, L. & Hazeldon, J. (1995) Salt marsh creationand management for coastal defence. In: Directions in European Coastal Management (Eds M.G. Healey& J.P. Doody), pp. 175–83. Samara Publishing,Cardigan.

Bornhold, B.D. & Giresse, P. (1985) Glauconitic sedi-ments on the continental shelf off Vancouver, BritishColumbia, Canada. Journal of Sedimentary Petrology55, 653–64.

Bosence, D.W.J. (1980) Sedimentary facies, produc-tion rates and facies models for recent coralline algalgravels, Co. Galway, Ireland. Geological Journal 15,91–111.

Bosence, D.W.J., Rowlands, R.J. & Quine, M.L. (1985)Sedimentology and budget of a Recent carbonatemound, Florida Keys. Sedimentology 32, 317–43.

Boss, S.K. & Liddell, D.W. (1987) Patterns of sedi-ment composition of Jamaican fringing reef facies.Sedimentology 34, 77–87.

Bothner, M.H., Parmenter, C.M. & Milliman, J.D.(1981a) Temporal and spatial variations in suspendedmatter in continental shelf and slope waters off thenorth-eastern United States. Estuarine, Coastal, andShelf Science 13, 213–34.

Bothner, M.H., Spiker, E.C., Johnson, P.P., et al. (1981b)Geochemical evidence for modern sediment accumula-tion on the continental shelf off south New England,Journal of Sedimentary Petrology 51, 281–92.

Bottino, G.M., Chiarle, M., Joly, A., et al. (2002) Model-ling rock avalanches and their relation to permafrostdegradation in glacier environments. Permafrost andPeriglacial Processes 13, 283–8.

Boult, S., Collins, D.N., White, K.N., et al. (1994) Metaltransport in stream polluted by acid mine drainage; theAfon Goch, Anglesey, UK. Environmental Pollution84, 279–84.

Bowen, H.J.M. (1966) Trace Elements in Biochemistry.Academic Press, London, 241 pp.

Boyd, N.A., Taylor, K.G. & Boult, S. (1999) Early diage-netic controls on contaminant geochemistry and mobil-ity in urban water bodies: Salford Quays, Manchester.Environmental Geochemistry and Health 22, 317–22.

Boyd, S.E., Limpenny, D.S., Rees, H.L., et al. (2003)Preliminary observations of the effects of dredgingintensity on the re-colonisation of dredged sedi-ments off the southeast coast of England (Area 222).Estuarine, Coastal and Shelf Science 56, 1–15.

Boyd, S.E., Cooper, K.M., Limpenny, D.S., et al. (2004)Assessment of the Re-habilitation of the SeabedFollowing Marine Aggregate Dredging. Science SeriesTechnical Report 121, Centre for Environment,Fisheries and Aquaculture Science, Lowestoft, 154 pp.(available online at www.cefas.co.uk).

Boyden, C.R., Aston, S.R. & Thornton, I. (1979) Tidaland seasonal variations of trace elements in two Cornishestuaries. Estuarine Coastal and Marine Science 9,303–17.

Bradbury, J.P. (1997) A diatom record of climate andhydrology for the past 200 ka from Owens Lake,California with comparison to other great basinrecords. Quaternary Science Reviews 16, 203–19.

Bradley, S.B. & Cox, J.J. (1990) The significance of thefloodplain to the cycling of metals in the river Derwentcatchment, UK. Science of the Total Environment97/98, 441–54.

Bradshaw, C. (1997) Bioturbation of reefal sediments by crustaceans in Phuket, Thailand. Proceedings ofEighth International Coral Reef Symposium, Vol. 2,pp. 1801–6. Smithsonian Tropical Research Institute,Bilbao, Republic of Panama.

Breed, C.S., McCauley, J.F., Whitney, M.I., et al. (1997)Wind erosion in drylands. In: Arid Zone Geomorpho-logy: Process, Form and Change in Drylands, 2nd edn(Ed. D.S.G. Thomas), pp. 437–64. Wiley, Chichester.

Bridge, J.S. (2003) Rivers and Floodplains. Forms, Pro-cesses, and Sedimentary Record. Blackwell Science,Oxford.

Bridgland, D.R. (2000) River terrace systems in north-west Europe: an archive of environmental change, uplift,and early human occupation. Quaternary ScienceReviews 19, 1293–303.

Bridgland, D.R. & Allen, P. (1996) A revised model forterrace formation and its significance for the earlyMiddle Pleistocene terrace aggradations of north-eastEssex, England. In: The Early Middle Pleistocene inEurope (Ed. C. Turner), pp. 121–34. Balkema,Rotterdam.

Bridgland, D.R. & Maddy, D. (2002) Global correlationof long Quaternary fluvial sequences: a review of baseline knowledge and possible methods and criteriafor establishing a database. Netherlands Journal ofGeosciences 81, 265–81.

Brinkhurst, R.O. (1974) The Benthos of Lakes.Macmillan, London, 189 pp.

Brinkmann, R. & Tobin, G.A. (2003) Urban SedimentRemoval: the Science, Policy, and Management ofStreet Sweeping. Kluwer Academic Press, Dordrecht,168 pp.

Bristow, C.S. (1996) Reconstructing channel morpho-logy from sedimentary sequences. In: Advances inFluvial Dynamics and Stratigraphy (Eds P.A. Carling& M.R. Dawson), pp. 351–71. Wiley, Chichester.

Brizga, S.O. & Finlayson, B.L. (1994) Interactionsbetween upland catchment and lowland rivers: anapplied Australian case study. Geomorphology 9(3),189–201.

Bromberg, S. (1990) Identifying ecological indicators: anenvironmental monitoring and assessment program.Journal of Air and Waste Management Association40, 976–87.

Bromhead, J.C. & Beckwith, P. (1994) Environmentaldredging on the Birmingham canals: water quality and sediment treatment. Journal of the Institution ofWater and Environmental Management 8, 350–9.

ESD01 9/5/06 3:09 PM Page 392

REFERENCES 393

Bromley, R.G. (1978) Bioerosion of Bermuda reefs.Palaeogeography, Palaeoclimatology, Palaeoecology23, 169–97.

Brookes, A. (1992) Recovery and restoration of someengineered British river channels. In: River Conserva-tion and Management (Eds P. Boon, P. Calow & G.E. Petts), pp. 337–53. Wiley, Chichester.

Brown, A.C. & McLachlan, A. (2002) Sandy shore eco-systems and the threats facing them: some predictionsfor the year 2025. Environmental Conservation 29,62–77.

Brown, A.G. (1997) Alluvial Geoarchaeology. FloodplainArchaeology and Environmental Change. CambridgeManuals in Archaeology, Cambridge University Press,Cambridge.

Brown, A.G. (1998) Fluvial evidence of the Medievalwarm period and the late Medieval climatic deteriora-tion in Europe. In: Palaeohydrology and EnvironmentalChange (Eds G. Benito, V.R. Baker & K.J. Gregory),pp. 43–52. Wiley, Chichester.

Brunet, R.C. & Astin, K.B. (2000) A 12-month sedimentand nutrient budget in a floodplain reach of the RiverAdour, southwest France. Regulated Rivers – Researchand Management 16, 267–77.

Brunsden, D. & Moore, R. (1999) Engineering geo-morphology on the coast: lessons from west Dorset.Geomorphology 31, 391–409.

Brunskill, G.J., Zagorskis, I. & Pfitzner, J. (2002)Carbon burial rates in sediments and a carbon massbalance for the Herbert River region of the GreatBarrier Reef continental shelf, north Queensland,Australia. Estuarine, Coastal and Shelf Science 54,677–700.

Bruun, P. (1962) Sea level rise as a cause of shore erosion. Journal of Waterways and Harbours Division,American Society of Civil Engineers 88, 117–30.

Bryan, G.W. & Langstone, W.J. (1992) Bioavailability,accumulation and effects of heavy metals in sedi-ments with special reference to United Kingdom estuaries – a review. Environmental Pollution 76(2),89–131.

Bryant, E.A. (1988) The effect of storms on StanwellPark, N.S.W. beach position, 1943–1980. MarineGeology 79, 171–87.

Bryce, S., Larcombe, P. & Ridd, P.V. (1998) The relativeimportance of landward-directed tidal sediment trans-port versus freshwater flood events in the NormanbyRiver estuary, Cape York Peninsula, Australia.Marine Geology 149, 55–78.

Buddemeier, R.W. (1997) Making light work of adapta-tion. Nature 388, 229–30.

Buddemeier, R.W., Maragos, J.E. & Knutson, D.W.(1974) Radiographic studies of reef coral skeletons:rates and patterns of coral growth. Journal of Experi-mental Marine Biology and Ecology 14, 179–200.

Burd, F. (1995) Managed Retreat: a Practical Guide.English Nature, Peterborough.

Burger, R.L., Fulthorpe, C.S. & Austin, J.A. (2001) LatePleistocene channel incisions in the southern Eel RiverBasin, northern California: implications for tectonicvs. eustatic influences on shelf sedimentation patterns.Marine Geology 177, 317–30.

Burkham, D.E. (1972) Channel Changes of the GilaRiver in Safford Valley, Arizona. Professional Paper655-J, US Geological Survey, Denver, CO.

Burns, K.A., Garrity, S.D. & Levings, S.C. (1993) Howmany years until mangrove ecosystems recover fromcatastrophic oil spills? Marine Pollution Bulletin 26,239–48.

Burton, G.A. (2002) Sediment quality criteria in usearound the world. Limnology 3, 65–75.

Burton, Jr., G.A. (Ed.) (1992) Sediment Toxicity Assess-ment. Lewis, Boca Raton, 457 pp.

Bush, D., Pilkey, O.H. & Neal, W.J. (1996) Living by theRules of the Sea. Duke University Press, Durham, NC.

Butcher, S.S., Charlson, R.J., Orians, G.H. & Wolfe,G.V. (Eds) (1992) Global Biogeochemical Cycles.Academic Press, London, 379 pp.

Butler, A.J., Depers, A.M., McKillup, S.C., et al. (1977)Distribution and sediments of mangrove forests in southAustralia. Transactions of Royal Society of SouthernAustralia 101, 35–44.

Butler, D.R., Walsh, S.J. & Malanson, G.P. (2003)Introduction to the special issue: mountain geomor-phology – integrating earth systems. Geomorphology55, 1–4.

Butterfield, G.R. (1991) Grain transport rates in steadyand unsteady turbulent airflows. Acta MechanicaSupplement 1, 97–122.

Cacchione, D.A., Wiberg, P.L., Lynch, J., et al. (1999)Estimates of suspended-sediment flux and bedformactivity on the inner portion of the Eel continentalshelf. Marine Geology 154, 83–97.

Caine, N. (1974) The geomorphic processes of the alpineenvironment. In: Arctic and Alpine Environments(Eds J.D. Ives & R.G. Barry), pp. 721–48. Methuen,London.

Caine, N. (1984) Elevational contrasts in conmtem-porary geomorphic activity in the Colorado FrontRange. Studia Geomorphologica Carpatho-Balcanica18, 5–31.

Cairns, Jr., J. & Pratt, J.R. (1987) Ecotoxicologicaleffect indices: a rapidly evolving system. Water Scienceand Technology 19, 1–12.

Cairns, Jr., J. (1981) Testing for Effects of Chemicals inEcosystems. National Academy Press, Washington,DC, 103 pp.

Calhoun, R.S., Fletcher, C.H. & Harney, J.N. (2002) Abudget of marine and terrigenous sediments, HanaleiBay, Kauai, Hawaiian Islands. Sedimentary Geology150, 61–87.

Cameron, W.M. & Pritchard, D.W. (1963) Estuaries. In: The Sea (Ed. M.N. Hill), pp. 306–24. Wiley, New York.

ESD01 9/5/06 3:09 PM Page 393

394 REFERENCES

Carannante, G., Esteban, M., Milliman, J.D., et al.(1988) Carbonate lithofacies as paleolatitude indica-tors: problems and limitations. Sedimentary Geology60, 333–46.

Carr, A.P. (1969) Size grading along a pebble beach;Chesil Beach, England. Journal of Sedimentary Petro-logy 39, 297–311.

Carter, J., Owens, P.N., Walling, D.E., et al. (2003)Fingerprinting suspended sediment sources in a largeurban river system. The Science of the Total Environ-ment 314–16, 513–34.

Carter, L. (1986) A budget for modern-Holocene sediment on the South Otago continental shelf. NewZealand Journal of Marine and Freshwater Research20, 665–76.

Carter, L. & Carter, R.M. (1986) Holocene evolution ofthe nearshore sand wedge, South Otago continentalshelf, New Zealand. New Zealand Journal of Geologyand Geophysics 29, 413–24.

Carter, R.M., Carter, L., Williams, J.J., et al. (1985)Modern and relict sedimentation on the South Otago continental shelf, New Zealand. New ZealandOceanographic Institute Memoir 93, 43 pp.

Carter, R.W.G. (1988) Coastal Environments. AcademicPress, London.

Carter, R.W.G. (1988) Coastal Environments. CambridgeUniversity Press, Cambridge.

Carter, R.W.G. (1991) Near-future sea level impacts oncoastal dune landscapes. Landscape Ecology 6, 29–39.

Carter, R.W.G. & Orford, J.D. (1993) The morpho-dynamics of coarse clastic beaches and barriers: a short- and long term perspective. Journal of CoastalResearch 15, 158–79.

Carter, R.W.G. & Wilson, P. (1993) Aeolian processesand deposits in northwest Ireland. In: The Dynamicsand Environmental Context of Aeolian SedimentarySystems (Ed. K. Pye), pp. 173–90. Special PublicationNo. 72, Geological Society Publishing House, Bath.

Carter, R.W.G. & Woodroffe, C.D. (1994) CoastalEvolution. Cambridge University Press, Cambridge.

Carter, R.W.G., Orford, J.D., Forbes, D.L., et al. (1987)Gravel barriers headlands and lagoons: an evolution-ary model. In: Coastal Sediments ’87, Proceedings of aSpecialty Conference on Advances in Understandingof Coastal Sediment Processes (Ed. N.C. Kraus), pp. 1776–92. American Society of Civil Engineers,Reston, VA.

Carter, R.W.G., Eastwood, D.A. & Bradshaw, P. (1991)Small-scale sediment removal from beaches in NorthernIreland: environmental impact, community perceptionand conservation management. Aquatic Conservation:Marine and Freshwater Ecosystems 2, 95–113.

CEFAS (2001) Contaminant Status of The North Sea.Technical Report TR_004, produced for StrategicEnvironmental Assessment – SEA2. UK Depart-ment of Trade and Industry, Strategic EnvironmentalAssessment – SEA2, August 2001, 101 pp.

CEFAS (2003) The Use of Action Levels in theAssessment of Dredged Material Placement at Sea andin Estuarine Areas under the Food and EnvironmentalProtection Act (II). Report for the Department forEnvironment, Food and Rural Affairs, Project CodeAE0258 (available from www.defra.gov.uk)

Celliers, L. & Schleyer, M.H. (2002) Coral bleaching onhigh-latitude marginal reefs at Sodwana Bay, SouthAfrica. Marine Pollution Bulletin 44, 1380–7.

Chadwick-Furman, N.E. (1996) Reef coral diversity andglobal change. Global Change Biology 2, 559–68.

Chapman, P.M. (1986) Sediment quality criteria from asediment quality triad: an example. EnvironmentalToxicology and Chemistry 5, 957–64.

Chapman, P.M. (1996) Presentation and interpreta-tion of Sediment Quality Triad data. Ecotoxicology5, 327–39.

Chappell, J. & Grindrod, J. (1984) Chenier plain formation in northern Australia. In: Coastal Geomor-phology in Australia (Ed. B.G. Thom), pp. 197–231.Academic Press, Sydney.

Chappell, J., Omura, A., Esat, T., et al. (1996) Recon-ciliation of late Quaternary sea levels derived fromcoral terraces at Huon Peninsula with deep sea oxygenisotope records. Earth and Planetary Science Letters141, 227–36.

Charlesworth, S.M. & Foster, I.D.L. (1999) Sedimentbudgets and metal fluxes in two contrasting urban lakesin Coventry, UK. Applied Geography 19, 199–210.

Charlesworth, S.M. & Lees, J.A. (1999) The distributionof heavy metals in deposited urban dusts and sediments,Coventry, England. Environmental Geochemistry andHealth 21, 97–115.

Charlesworth, S.M. & Lees, J.A. (2001) The applicationof some mineral magnetic measurements and heavymetal analysis for characterising fine sediments in anurban catchment, Coventry, UK. Journal of AppliedGeophysics 48, 113–25.

Charlesworth, S.M., Everett, M., McCarthy, R., et al.(2003a) A comparative study of heavy metal distribu-tion in deposited street dusts in a large and small urbanarea: Birmingham and Coventry, West Midlands, UK.Environment International 29, 563–73.

Charlesworth, S.M., Harker, E. & Rickard, S. (2003b)Review of sustainable drainage systems (SuDS): a softoption for hard drainage questions? Geography 88,99–107.

Chatres, C.J., Greene, R.S., Ford, G.W., et al. (1985)The effects of gypsum on macroporosity and crustingof two duplex soils. Australian Journal of Soil Research23, 467–79.

Chen, S., Billings, S.A. & Grant, P.M. (1990) Non-linearsystem identification using neural networks. Inter-national Journal of Control 51, 1191–214.

Chen, Y.X., Zhu, G.W., Tian, G.M., et al. (2002) Phyto-xicity of dredged sediment from urban canal as landapplication. Environmental Pollution 117, 233–41.

ESD01 9/5/06 3:09 PM Page 394

REFERENCES 395

Chesnokov, A.V., Govorun, A.P., Linnik, V.G., et al.(2000) 137Cs contamination of the Techa flood plainnear the village of Muslumovo. Journal of Environ-mental Radioactivity 50, 179–91.

Chmura, G.L., Coffrey, A. & Crago, R. (2001) Variation insurface sediment deposition on salt marshes in the Bayof Fundy. Journal of Coastal Research 17(1), 221–7.

Choy, S.C. & Booth, W.E. (1994) Prolonged inundationand ecological changes in an Avicennia mangrove:implications for conservation and management. Hydro-biologia 285, 237–47.

Church, M. & Ryder, J.M. (1972) Paraglacial sedimenta-tion: a consideration of fluvial processes conditionedby glaciation. Geological Society of America Bulletin83, 3059–71.

Church, M. & Slaymaker, O. (1989) Disequilibrium ofHolocene sediment yield in glaciated British Columbia.Nature 337, 452–54.

Churchill, J.H. (1989) The effect of commercial trawlingon sediment resuspension and transport over the middle Atlantic Bight continental shelf. ContinentalShelf Research 9, 841–64.

Clark, C.F., Smith, P.G., Nielson, G., et al. (2000)Chemical characterisation and legal classification ofsludges from roads sweepings. Journal of the CharteredInstitution of Water and Environmental Management14, 99–102.

Clark, M.W. (1998) Management implications of metaltransfer pathways from refuse tip to mangrove sediments.The Science of the Total Environment 222, 17–34.

Clark, R.B. (1992) Marine Pollution, 3rd edn. ClarendonPress, Oxford.

Clarke, D.J. & Eliot, I.G. (1988) Low frequency changesof sediment volume on the beachface at Warilla beach,New South Wales, 1975–1985. Marine Geology 79,189–211.

Clavier, J., Chardy, P. & Chevillon, C. (1995) Sedi-mentation of particulate matter in the south-westlagoon of New Caledonia. Estuarine, Shelf and CoastalScience 40, 281–94.

Cleary, W.J., Hosier, P.E. & Wells, J. (1976) Genesisand significance of marsh islands in southeastern NorthCarolina lagoons. Journal of Sedimentary Petrology49, 703–10.

Clements, T., Merriam, R.H., Stone, R.O., et al. (1957)A Study of Desert Conditions. Technical Report EP-53, Headquarters Quartermaster Research andDevelopment Command, US Army EnvironmentalProtection Research Division, Natck, MA.

Clough, B.F., Boto, K.G. & Attiwell, P.M. (1983) Man-groves and sewage: a re-evaluation. In: Tasks forVegetation Science, Vol. 8 (Ed. H.J. Teas), pp. 151–61.Dr W. Junk Publishers, The Hague.

Cochran, J.K., Moran, S.B., Fisher, N.S., et al. (2000)Sources and transport of anthropogenic radionuclidesin the Ob River system, Siberia. Earth and PlanetaryScience Letters 179, 125–37.

Cohen, J.E., C. Small, C., Mellinger, A., et al. (1997)Estimates of coastal populations Science 278, 1211–12.

Cole, K.L. & Mayer, L. (1982) Use of packrat middensto determine rates of cliff retreat in the Eastern GrandCanyon, Arizona. Geology 10, 597–9.

Coleman, J.M. (1988) Dynamic changes and processesin the Mississippi river delta. Geological Society ofAmerica Bulletin 100, 999–1015.

Coleman, M.L. (1985) Geochemistry of diagenetic non-silicate minerals: kinetic considerations. PhilosophicalTransactions of the Royal Society of London Series A315, 39–56.

Collins, A.L. & Walling, D.E. (2002) Selecting finger-print properties for discriminating potential suspendedsediment sources in river basins. Journal of Hydrology261, 218–44.

Collins, A.L., Walling, D.E. & Leeds, G.J.L. (1997)Source type ascription for fluvial suspended sedimentbased on a quantitative composite fingerprinting tech-nique. Catena 29, 1–17.

Collins, B.D. & Dunne, T. (1986) Erosion of tephrafrom the 1980 eruption of Mount St Helens. Geo-logical Society of America Bulletin 97, 896–905.

Collins, M. & Evans, G. (1986) The influence of fluvialsediment supply on coastal erosion in west and central Africa. Journal of Shoreline Management 2, 5–12.

Collinson, J.D. (1986) Alluvial sediments. In: SedimentaryEnvironments and Facies, 2nd edn (Ed. H.G. Reading),pp. 20–62. Blackwell Scientific, Oxford.

Colman, S.M., King, J.W., Jones, G.A., et al. (2000)Holocene and recent sediment accumulation rates insouthern Lake Michigan. Quaternary Science Reviews19, 1563–80.

Coltorti, M. (1997) Human impact in the Holocenefluvial and coastal evolution of the Marche region,Italy. Catena 30, 311–35.

Cooke, R.U. & Doornkamp, J.C. (1990) Geomorphologyin Environmental Management: a New Introduction,2nd edn. Clarendon Press, Oxford.

Cooke, R.U., Brunsden, D., Doornkamp, J.C., et al.(Eds) (1982) Urban Geomorphology in Drylands.Oxford University Press, Oxford, 324 pp.

Cooke, R.U., Warren, A. & Goudie, A.S. (1993) DesertGeomorphology. University College London Press,London, 526 pp.

Cooper, D.C. & Morse, J.W. (1998) Extractability ofmetal sulphide minerals in acidic solutions: applicationsto environmental studies of trace metal contaminationwithin anoxic sediments. Environmental Science andTechnolology 32, 1076–8.

Cooper, J.A.G. (1991) Beachrock formation in low latitudes: implications for coastal evolutionary models.Marine Geology 98, 145–54.

Cooper, J.A.G. (1993) Sedimentary processes in theriver-dominated Mvoti estuary, South Africa. Geo-morphology 9, 271–300.

ESD01 9/5/06 3:09 PM Page 395

396 REFERENCES

Cooper, J.A.G. (2001a) Geomorphology of tide-dominated and river-dominated, barred microtidalestuaries: a contrast. Journal of Coastal ResearchSpecial Issue 34, 428–36.

Cooper, J.A.G. (2001b) Gemorphological variabilityamong microtidal estuaries from the wave-dominatedSouth African coast. Geomorphology 40, 99–122.

Cooper, J.A.G. & Navas, F. (2004) Natural bathymetricchange as a control on century-scale shoreline beha-vior. Geology 32, 513–16.

Cooper, J.A.G. & Pilkey, O.H. (2004a) Alternatives tothe mathematical modelling of beaches. Journal ofCoastal Research 20, 641–4.

Cooper, J.A.G. & Pilkey, O.H. (2004b) Longshore drift:trapped in an expected universe. Journal of Sediment-ary Research 74, 1–15.

Cooper, J.A.G., Wright, C.I. & Mason, T.R. (1999)Geomorphology and sedimentology of South Africanestuaries. In: Estuaries of Southern Africa (Eds B.R. Allanson & D. Baird), pp. 5–25. CambridgeUniversity Press, Cambridge.

Cooper, N.J. & Jay, H. (2002) Predictions of large-scalecoastal tendency: development and application of aqualitative behaviour-based methodology. Journal ofCoastal Research, Special Issue 36, 173–81.

Cordery, I., Pilgrim, D.H. & Doran, D.G. (1983) Somehydrological characteristics of arid western NewSouth Wales. In: Hydrology and Water ResourcesSymposium 1983, Institute of Australian Engineers,National Conference Publication 83/13, pp. 287–92.

Costa, J.E. (1987) Hydraulics and basin morphometryof the largest flash floods in the conterminous UnitedStates. Journal of Hydrology 93, 313–38.

Costa, J.E. (1988) Rheologic, geomorphic, and sedimento-logic differentiation of water floods, hyperconcentratedflows and debris flows. In: Flood Geomorphology (EdsV.R. Baker, R.C. Kochel & P.C. Patton), pp. 113–22.Wiley, New York.

Coulthard, T.J. & Macklin, M.G. (2001) How sensitiveare river systems to climate and land-use changes? A model-based evaluation. Journal of QuaternaryScience 16, 347–51.

Coulthard, T.J., Kirkby, M.J. & Macklin, M.G. (2000)Modelling geomorphic response to environmentalchange in an upland catchment. Hydrological Processes14, 2031–45.

Cousset, P. & Meunier, M. (1996) Recognition, clas-sification and mechanical description of debris flows.Earth Science Reviews 40, 209–27.

Cowell, P.J., Roy, P.S. & Jones, R.A. (1995) Simulationof large-scale coastal behaviour using a morphologicalbehaviour model. Marine Geology 126, 45–61.

Cronin, S.J., Neall, V.E., Lecointre, J.A., et al. (1997)Changes in Whangauhu river lahar characteristicsduring the 1995 eruption sequence, Ruapehu volcano,New Zealand. Journal of Volcanology and GeothermalResearch 76, 47–61.

Crooks, S., Schutten, J., Sheern, G.D., et al. (2002)Drainage and elevation as factors in the restoration of salt marshes in Britain. Restoration Ecology 10(3),591–602.

Cummings, K.W. (1973) Trophic relations in aquaticinsects. Annual Reviews in Entomology 18, 183–206.

Cundy, A.B., Croudace, I.W., Thomson, J., et al. (1997)Reliability of salt marshes as ‘geochemical recorders’of pollution input: a case study from contrasting estuaries in southern England. Environmental Scienceand Technology 31, 1093–101.

Cunha, P.P., Martins, A.A., Daveau, S., et al. (2005)Tectonic control of the Tejo river fluvial incision during the late Cenozoic, in Ródão – central Portugal(Atlantic Iberian border). Geomorphology 64, 271–98.

Daan, R. & Mulder, M. (1996) On the short-term andlong-term impact of drilling activities in the Dutch sector of the North Sea. – ICES Journal of MarineScience 53, 1036–44.

Dando, P.R. (2001) A review of pockmarks in the UKpart of the North Sea, with particular respect to theirbiology. Technical Report 001 – Pockmarks. StrategicEnvironmental Assessment – SEA2. UK Departmentof Trade and Industry, 22 pp.

Darmody, R.G. & Foss, J.E. (1979) Soil-landscape rela-tions of the tidal marshlands of Maryland. Journal ofthe Soil Science Society of America. 43, 534–41.

Davidson, N.C., d’A Laffoley, D., Doody, J.P., et al.(1991) Nature Conservation and Estuaries in GreatBritain. Nature Conservancy Council, Peterborough.

Davies, A.G. & Villaret, C. (2002) Prediction of sandtransport rates by waves and currents in the coastalzone. Continental Shelf Research 22, 2725–37.

Davies, B.E. & Lewin, J. (1974) Chronosequences in alluvial soils with special reference to historical pollution in Cardiganshire, Wales. EnvironmentalPollution 6, 49–57.

Davies, J.H. (1964) A morphogenetic approach to worldshorelines. Zeitscrift fur Geomorphologie 8, 127–42.

Davies, J.L. (1980) Geographical Variation in CoastalDevelopment. Oliver and Boyd, Edinburgh.

Davies, T.R. (1997) Long-term management of facilitieson an active alluvial fan – Waiho River fan, Westland,New Zealand. Journal of Hydrology (New Zealand)36, 127–45.

Davies, T.R. & McSaveney, M.J. (2001) Anthropogenicfanhead aggradation, Waiho River, Westland, NewZealand. In: Gravel-bed Rivers V (Ed. M.P. Mosley),pp. 531–53. The Caxton Press, Christchurch.

Davies, T.R., McSaveney, M.J. & Clarkson, P.J. (2003)Anthropic aggradation of the Waiho River, Westland,New Zealand: microscale modelling. Earth SurfaceProcesses and Landfroms 28, 209–18.

Davies, T.R.H. (1991) Research of fluvial processes in mountains – a change of emphasis. In: FluvialHydraulics of Mountain Regions (Eds A. Armanini &G. Di Silvio), pp. 251–66. Springer-Verlag, Berlin.

ESD01 9/5/06 3:09 PM Page 396

REFERENCES 397

Davis, J.A. & Leckie, J.O. (1978) Effects of adsorbeccomplexing ligands on trace metal uptake by hydrousoxides. American Chemical Society 12, 1309–15.

Davis, R.A. & Hayes, M.O. (1984) What is a wave-dominated coast? Marine Geology 60, 313–29.

Davis, R.A. Jr., Welty, A.T., Borrego, J., et al. (2000) Rio Tinto estuary (Spain): 5000 years of pollution.Environmental Geology 39, 1107–16.

Dawson, A.G. (1994) Geomorphological processes associated with tsunami run-up and backwash. Geo-morphology 10, 83–94.

De Miguel, E., Llamas, J.F., Chacon, E., et al. (1997)Origin and patterns of distribution of trace elements in street dust: unleaded petrol and urban lead.Atmospheric Environment 31, 2733–40.

De Ruig, J.H.M. & Hillen, R. (1997) Developments inDutch coastline management: conclusions from thesecond governmental coastal report. Journal of CoastalConservation 3, 203–10.

De Sylva, D. (1986) Increased storms and estuarinesalinity and other ecological impacts of the green-house effect. In: Effects of Changes in StratosphericOzone and Global Climate, Vol. 4 (Ed. J.G. Titus),pp. 153–94. U.S. Environmental Protection Agency,Washington.

Dean, R.G. (1974) Compatibility of borrow material forbeach fills. Proceedings of the 14th Coastal Engineer-ing Conference, New York, American Society of CivilEngineers, pp. 1319–33.

Debenay, J.-P. (1987) Sedimentology in the south-western lagoon of New Caledonia, SW Pacific. Journalof Coastal Research 3, 77–91.

Debenay, J.P., Guiral, D. & Parra, M. (2002) Eco-logical factors acting on the microfauna in mangroveswamps. The case of foraminiferal assemblages inFrench Guiana. Estuarine, Coastal and Shelf Science55, 509–33.

Dedkov, A.P. & Moszherin, V.I. (1992) Erosion andsediment yield in mountain regions of the world. In: Erosion, Flows and Environment in MountainRegions. Proceedings of the Chengdo Symposium,1992 (Eds D.E. Walling, T.R. Davies & B. Hasholt),pp. 29–36. IAHS Publication 209, InternationalAssociation of Hydrological Sciences, Wallingford.

DEFRA (2005) Charting Progress: an Integrated Assess-ment of the State of UK Seas. Department for Envir-onment, Food and Rural Affairs, London, 130 pp.(www.defra.gov.uk).

Degens, E.T., Kempe, S. & Richey, J.E. (1991)Summary: Biogeochemistry of major world rivers. In:Biogeochemistry of Major World Rivers, SCOPE 42(Eds E.T. Degens, S. Kempe & J.E. Richey), pp. 323–47. Wiley, Chichester.

DeLaune, R.D., Smith, C.J., Patrick, W.H., et al. (1994) Effect of oil on salt marsh biota: methods forrestoration. Environmental Pollution, Series A. 36,207–27.

Dennis, I.A., Macklin, M.G., Coulthard, T.J., et al.(2003) The impact of the October–November 2000floods on contaminant metal dispersal in the RiverSwale catchment, North Yorkshire, UK. HydrologicalProcesses 17, 1641–57.

Department of Trade and Industry (2001) StrategicEnvironmental Assessment of the Mature Areas of theOffshore North Sea (SEA 2). Consultation Document.September 2001, 235 pp. (available from www.habitats-directive.org).

Descroix, L. & Gautier, E. (2002) Water erosion in thesouthern French alps: climatic and human mechanisms.Catena 50, 53–85.

Desprez, M. (2000) Physical and biological impact ofmarine aggregate extraction along the French coast of the Eastern English Channel: short- and long-termpost-dredging restoration. ICES Journal of MarineScience 57, 1428–38.

Dietrich, W.E., Kirchner, J.W., Ikeda, H., et al. (1989)Sediment supply and the development of the coarsesurface layer in gravel-bedded rivers. Nature 340,215–17.

Diskin, M.H. & Lane, L.J. (1972) A basinwide stochasticmodel of ephemeral stream runoff on southeasternArizona. International Association Scientific Hydro-logists Bulletin 17, 61–76.

Dittmar, T., Lara, R.J. & Kattner, G. (2001) River ormangrove? Tracing major organic matter sources intropical Brazilian coastal waters. Marine Chemistry73, 253–71.

Dodd, J., Large, D.J., Fortey, N.J., et al. (2003)Geochemistry and petrography of phosphorus inurban canal bed sediment. Applied Geochemistry18, 259–67.

DOER (2005) Dredging Operations and EnvironmentalResearch, U.S. Army Corps of Engineers, Vicksburg,Mississippi (http://el.erdc.usace.army.mil/dots/doer/).

Dominguez, J.M.L., Martin, L. & Bittencourt, A.C.S.P.(1987) Sea-level history and Quaternary evolution of river mouth-associated beach-ridge plains along theeast-southeast Brazilian coast: a summary. In: Sea-levelFluctuation and Coastal Evolution (Eds D. Nummedal,O.H. Pilkey & J.D. Howards), pp. 115–27. SpecialPublication 41, Society of Economic Paleontologistsand Mineralogists, Tulsa, OK.

Done, T.J. (1992) Phase shifts in coral reef communitiesand their ecological significance. Hydrobiologia 247,121–32.

Doornkamp, J.C. & Ibrahim, H.A.M. (1990) Salt weather-ing. Progress in Physical Geography X, 335–48.

Douillet, P. (1998) Tidal dynamics of the south-westlagoon of New Caledonia: observations and 2D numer-ical modelling. Oceanologica Acta 21, 69–79.

Douillet, P., Ouillon, S. & Cordier, E.A. (2001) A numer-ical model for fine suspended sediment transport in the south-west lagoon of New Caledonia. Coral Reefs20, 361–72.

ESD01 9/5/06 3:09 PM Page 397

398 REFERENCES

Dregne, H.E. (1976) Soils of Arid Regions. Elsevier,Amsterdam, 237 pp.

Dregne, H.E. (1983) Desertification of Arid Lands.Harwood Academic, Chur, Switzerland.

Driedger, C.L. (1988) Geology in action – jökulhlaupson Mount Rainier. Water Fact Sheet, Open File Report88-459, US Geological Survey, Denver, CO.

Driedger, C.L. & Fountain, A.G. (1989) Glacier outburstfloods at Mount Rainier, Washington State, U.S.A.Annals of Glaciology 13, 51–5.

Droppo, I.G., Irvine, K.N., Murphy, T.P., et al. (1998)Fractionated metals in street dust of a mixed land-usesewershed, Hamilton, Ontario. In: Hydrology in aChanging Environment, Vol III (Eds H. Wheater & C. Kirby), pp. 383–94. Wiley, New York.

Dubko, N.V. (1985) The Labile and Stable OrganicMatter. Ecological System of Naroch Lakes, Minsk,pp. 233–7. (In Russian.)

Duggan, M.J. & Williams, S. (1977) Lead-in-dust in city streets. The Science of the Total Environment 7,91–7.

Duke, N.C. & Watkinson, A.J. (2002) Chlorophyll-deficient propagules of Avicennia marina and appar-ent longer term deterioration of mangrove fitness inoil-polluted sediments. Marine Pollution Bulletin 44,1269–76.

Duke, N.C., Burns, K.A., Swannell, R.P.J., et al. (2000)Dispersant use and a bioremediation strategy as alternate means of reducing impacts of large oil spills on mangroves: the Gladstone field trial. MarinePollution Bulletin 41, 403–12.

Dunkerley, D.L. & Brown, K.J. (1997) Desert Soils. In:Arid Zone Geomorphology: Process, Form and Changein Drylands, 2nd edn (Ed. D.S.G. Thomas), pp. 55–68.Wiley, Chichester.

Dunne, T., Bren, D., Mertes, L.A.K., et al. (1988)Exchanges of sediment between the flood plain andchannel of the Amazon River in Brazil. GeologicalSociety of America Bulletin 110, 450–67.

Duplisea, D.E., Jennings, S., Malcolm, S.J., et al. (2001)Modelling potential impacts of bottom trawl fisherieson soft sediment biogeochemistry in the North Sea.Geochemical Transactions 14, 1–6.

Dye, K.R. (1998) The typology of intertidal mudflats. In: Sedimentary Processes in the Intertidal Zone (EdsK.S. Black, D.H. Paterson & A. Cramp), pp. 11–24.Special Publication 139, Geological Society PublishingHouse, Bath.

Dyer, K.R. (1986) Coastal and Estuarine SedimentDynamics. Wiley Interscience. New York, 342 pp.

Dyer, K.R. (1997) Estuaries: a Physical Introduction,2nd edn. Wiley, Chichester.

Dyer, K.R. (2000) Estuaries, A Physical Introduction,2nd edn. Wiley.

Dyrynda, P. (1996) Barrages within estuaries – ecologicallessons from the Tawe development. Marine Update 25.World Wildlife Fund, Godalming.

Eakin, C.M. (1996) Where have all the carbonates gone?A model comparison of calcium carbonate budgetsbefore and after the 1982–1983 El Niño at Uva Islandin the eastern Pacific. Coral Reefs 15, 109–19.

Eberhardt, E., Willenberg, H., Loew, S., et al. (2001)Active rockslides in Switzerland – understandingmechanisms and processes. In: International Confer-ence on Landslides – Causes, Impacts and Counter-measures (Eds M. Kü, H.H. Einstein, E. Krauter, H. Klapperich & R. Pöttler) pp. 25–34. 17–21 June,Davos, Switzerland, VGE, Essen.

Eden, D.E. & Page, M.J. (1998) Paleoclimatic implica-tions of a storm erosion record from late Holocencelake sediments, North Island, New Zealand. Palaeo-geography, Palaeoclimatology, Palaeoecology 139,37–58.

Edinger, E.N., Limmon, G.V. & Jompa, J. (2000)Normal coral growth rates on dying reefs: are coralgrowth rates good indictors of reef health. MarinePollution Bulletin 40, 404–25.

Edwards, A.J. & Clark, S. (1998) Coral transplantation:a useful management tool or misguided meddling?Marine Pollution Bulletin 37, 474–87.

Ehrlich, P.R. & Ehrlich, A.H. (1970) Population,Resources, Environment, 2nd edn. Freeman, SanFrancisco, 383 pp.

Einstein, H.A. (1950) The Bed-load Function for Sedi-ment Transportation in Open-channel Flows. TechnicalBulletin 1026, U.S. Department of Agriculture andSoil Conservation Service, Washington, DC.

Eisbacher, G.H. & Clague, J.J. (1984) Destructive MassMovements in High Mountains: Hazard and Manage-ment. Paper 84-16, Geological Survey of Canada,Vancouver, British Columbia.

Elliott, M. & Cutts, N.D. (2004) Marine habitats: loss and gain, mitigation and compensation. MarinePollution Bulletin 49, 671–74.

Ellison, A.M. & Farnsworth, E.J. (1996) Anthropogenicdisturbance of Caribbean mangrove ecosystems: past impacts, present trends, and future predictions.Biotropica 28, 549–65.

Ellison, J.C. & Stoddart, D.R. (1991) Mangrove ecosys-tem collapse during predicted sea-level rise: Holoceneanalogues and implications. Journal of Coastal Research7, 151–65.

Ellison, J.C. (1998) Impacts of sediment burial on mangroves. Marine Pollution Bulletin 37, 420–6.

Emergy, K.O. & Kuhn, G.G. (1982) Sea cliffs: their pro-cesses, profiles and classification. Geological Societyof America Bulletin 93, 644–54.

Emmerson, R.H.C., Birkett, J.W., Scromshaw, M., et al.(2000) Solid phase partitioning of metals in managedretreat soils: field changes over the first year of tidalinundation. The Science of the Total Environment254, 75–92.

EMU (2004) Marine Aggregate Site Restoration andEnhancement: a Strategic Feasibility and Policy Review.

ESD01 9/5/06 3:09 PM Page 398

REFERENCES 399

Report to the British Marine Aggregate ProducersAssociation, The Crown Estate and English Nature,125 pp. (available from www.bmapa.org.uk).

Environment Agency UK (2004) Foresight Future Flood-ing Report. http://www.environment-agency.gov.uk/subjects/flood/763964/763974/?version=18lang=_e(accessed 25 June 2004).

Esslemont, G. (2000) Heavy metals in seawater, marinesediments and corals from the Townsville section,Great Barrier Reef Marine Park, Queensland. MarineChemistry 71, 215–31.

Eswaran, H., Reich, P. & Beinroth, F. (2001) Globaldesertification tension zones. In: Sustaining the Global Farm (Eds D.E. Stott, R.H. Mohtar & G.C. Steinhardt), pp. 24–8. Selected Papers, 10thInternational Soil Convention Organization Confer-ence, 24–29 May 1999, Purdue. National Soil ErosionResearch Laboratory, Agriculture Research Service,US Department of Agriculture. Purdue University,West Lafayette, Indiana.

European Marine Strategy (2005) Proposal for a Direc-tive of the European Parliament and of the Councilestablishing a Framework for Community Action inthe field of Marine Environmental Policy (MarineStrategy Directive). 2005/0211 (COD), EuropeanUnion (available at http://europa.eu.int).

Evans, J.K., Gottgens, J.F., Gill, W.M. & Mackey, S.D.(2000) Sediment yields controlled by intrabasinal storage and sediment conveyance over the interval1984–1994. Journal of Soil and Water Conservation55, 264–70.

Evans, M. & Slaymaker, O. (2004) Spatial and temporalvariability of sediment delivery from alpine lake basins,Cathedral Provincial Park, southern British Columbia.Geomorphology 61, 209–24.

Evenari, M., Shanan, L. & Tadmor, N. (1982) The Negev:the Challenge of a Desert, 2nd edn. Harvard Univer-sity Press, Cambridge, MA, 437 pp.

Eyre, B. & McConchie, D. (1993) Implications of sedimentological studies for environmental pollutionassessment and management: example from fluvialsystems in north Queensland and western Australia.Sedimentary Geology 85, 235–52.

Fallon, S.J., White, J.C. & McCulloch, M.T. (2002)Porites corals as recorders of mining and environ-mental impacts: Misima Island, Papua New Guinea.Geochimica et Cosmochimica Acta 66, 45–62.

Farago, M.E., Kavanagh, P., Blanks, R., et al. (1998)Platinum concentrations in urban road dust and soil,and in blood and urine in the United Kingdom.Analyst 123, 451–4.

Farella, N., Lucotte, M., Louchouarn, P., et al. (2001)Deforestation modifying terrestrial organic trans-port in the Rio Tapajos, Brazilian Amazon. OrganicGeochemistry 32, 1443–58.

Farmer, A. (1997) Managing Environmental Pollution.Routledge, London, 246 pp.

Farmer, J.F. & Lyon, T.D.B. (1977) Lead in Glasgowstreet dirt and soil. Science of the Total Environment8, 89–93.

Farquarson, F.A.K., Meigh, J.R. & Sutcliffe, J.V. (1992)Regional flood frequency analysis in arid and semi-arid areas. Journal of Hydrology 138, 487–501.

Fergusson, J.E. & Kim, N.D. (1991) Trace elements in street and house dusts: sources and speciation. TheScience of the Total Environment 100, 125–50.

Field, C.D. (1998) Rehabilitation of mangrove eco-systems: an overview. Marine Pollution Bulletin 37,383–92.

Figueiredo, A.G. Jr., Nittrouer, C.A., Faria, L.E.C. Jr., et al. (1996) Gas-charged sediments in the Amazonsubmarine delta. Geomarine Letters 16, 31–5.

Filipek, L.H., Nordstrom, D.K. & Ficklin, W.H. (1987)Interaction of acid-mine drainage with waters andsediments of West Squaw Creek in the West Shastamining district, California. Environmental Scienceand Technology 21, 388–96.

Firth, C.R., Smith, D.E., Hansom, J.D., et al. (1995)Holocene spit development on a regressive shore-line, Dornoch Firth, Scotland. Marine Geology 124,203–14.

Foley, M.G. (1978) Scour and fill in steep, sand-bedephemeral streams. Geological Society of AmericaBulletin 89, 559–70.

Fookes, P.G., Sweeney, H., Manby, C.N.D., et al. (1985)Geological and geotechnical engineering aspects oflow cost roads in mountainous terrain. EngineeringGeology 21, 1–152.

Ford, T.D. & Pedley, H.M. (1996) A review of tufa andtravertine deposits of the world. Earth-ScienceReviews 41, 117–75.

Forristall, G.Z., Hamilton, R.C. & Cardone, V.J. (1977)Continental shelf currents in Tropical Storm Delia:observations and theory. Journal of Physical Oceano-graphy 7, 532–46.

Förstner, U. & Salomons, W. (1981) Trace MetalAnalysis in Polluted Sediments. Publication No. 248,Delft Hydraulics Laboratory, pp. 1–13.

Förstner, U. & Müller, G. (1974) Schwermetalle in Flüssenund Seen. Springer-Verlag, Berlin, 225 pp.

Förstner, U. & Wittmann, G.T.W. (1981) Metal Pollu-tion in the Aquatic Environment. Springer-Verlag,Berlin, 486 pp.

Foster, I.D.L. & Walling, D.E. (1994) Using reservoirdeposits to reconstruct changing sediment yields andsources in the catchment of the Old mill Reservoir,South Devon, UK, over the past 50 years. Hydro-logical Sciences Journal 39, 347–68.

Fournier, F. (1960) Climat et Érosion: La Relation Entre l’Érosion du Sol par l’Eau et les PrécipitationsAtmosphériques. Presses Universitaires France, Paris.

Francingues, N.R., Palermo, M.R., Lee, C.R., et al. (1985)Management Strategy for Disposal of Dredged Material:Contaminant Testing and Controls. Miscellaneous

ESD01 9/5/06 3:09 PM Page 399

400 REFERENCES

Paper D-85-1, U.S. Army Engineer Waterways Experi-ment Station, Vicksburg, MS.

Fraser, R.J. (1993) Removing contaminated sedimentsfrom the coastal environment: the New BedfordHarbour project example. Coastal Management 21,155 –62.

French, P.W. (1990) Coal dust: a marker pollutant in the Severn Estuary and Bristol Channel. UnpublishedPhD thesis. Postgraduate Research Institute for Sedi-mentology, University of Reading.

French, P.W. (1996) Long-term variability of copper,lead and zinc in salt marsh sediments of the SevernEstuary, UK. Mangroves and Salt Marshes 1(1),59–68.

French, P.W. (1997) Coastal and Estuarine Management.Routledge, London.

French, P.W. (2001) Coastal Defences: Processes, Prob-lems and Solutions. Routledge, London.

Frey, R.W. & Basan, P.B. (1978) Coastal salt marshes.In: Coastal Sedimentary Environments (Ed. R.A. Davis),pp. 101–69. Springer-Verlag, New York.

Friedman, G.M. & Sanders, J.E. (1978) Principles ofSedimentology. Wiley, New York, 792 pp.

Froelich, P.N., Klinkhammer, G.P., Bender, M.L., et al.(1979) Early oxidation of organic matter in pelagicsediments of the eastern equatorial Atlantic: suboxicdiagenesis. Geochimica et Cosmochimica Acta 43,1075–90.

Froidefrond, J.M., Pujos, M. & Andre, X. (1988)Migration of mud banks and changing coastline inFrench Guiana. Marine Geology 84, 19–30.

Frostick, L.E. & Reid, I. (Eds) (1987) Desert Sediments:Ancient and Modern. Geological Society of LondonSpecial Publication 35. Blackwell Scientific, Oxford,401 pp.

Frostick, L.E., Reid, I. & Layman, J.T. (1983) Changingsize distribution of suspended sediment in arid-zoneflash floods. In: Modern and Ancient Fluvial Systems(Eds J.D. Collinson & J. Lewin), pp. 97–106. SpecialPublication 6, International Association of Sedimento-logists. Blackwell Scientific, Oxford.

Fryberger, S.G. & Goudie, A.S. (1981) Arid Geomor-phology. Progress in Physical Geography 5, 420–8.

Fryberger, S.G., Schenk, C.J. & Krystinik, L.F. (1988)Stokes surfaces and the effects of near surface ground-water-table on aeolian deposition. Sedimentology 35,21–41.

Furakawa, K. & Wolanski, E. (1996) Sedimentation inmangrove forests. Mangroves and Salt Marshes 1,3–10.

Furukawa, K., Wolanski, E. & Mueller, H. (1997)Currents and sediment transport in mangrove forests.Estuarine, Coastal and Shelf Science 44, 301–10.

Fütterer, D.K. (1974) Significance of the boring spongeCliona for the origin of fine grained material of car-bonate sediments. Journal of Sedimentary Petrology44, 79–84.

Gaeuman, D., Schmidt, J.C. & Wilcock, P.R. (2005)Complex channel responses to changes in stream flowand sediment supply on the lower Duschesne River,Utah. Geomorphology 64, 185–206.

Gagan, M.K., Chivas, A.R. & Herczeg, A.L. (1990)Shelf-wide erosion, deposition, and suspended sedi-ment transport during Cyclone Winifred, centralGreat Barrier Reef, Australia. Journal of SedimentaryPetrology 60, 456–70.

Gagan, M.K., Johnson, D.P. & Carter, R.M. (1988) TheCyclone Winifred storm bed, central Great BarrierReef shelf, Australia. Journal of Sedimentary Petrology58, 845–56.

Gaines, R.A. & Maynord, S.T. (2001) Microscale loose-bed hydraulic models. Journal of HydraulicEngineering May, 335–8.

Gartner, J.W., Cheng, R.T., Wang, P-F., et al. (2001)Laboratory and field evaluations of the LISST-100instrument for suspended particle size determinations.Marine Geology 175, 199–219.

Gerard, C. & Chocat, B. (1998) An aid for the diagnosisof sewerage networks: analysis and modelling of thelinks between network’s physical structure and the riskof sediment build-up. Water Science and Technology39, 185–92.

German, J. & Svensson, G. (2002) Metal content and particle size distribution of street sediments and street sweeping waste. Water Science and Technology46, 191–8.

Gerrard, A.J. (1990) Mountain Environments. BelhavenPress, London.

GESAMP. (1982) The Review of the Health of theOceans. UNESCO, Paris.

Gibbs, R.J. (1973) Mechanisms of trace metal transportin rivers. Science 180, 71–3.

Gibbs, R.J. (1977) Transport phases of transition metalsin the Amazon and Yukon Rivers. Geological Societyof America Bulletin 88, 829–43.

Gilbertson, D.D. (1981) The impact of past and presentland use on a major coastal barrier system. AppliedGeography 1, 97–119.

Gilbertson, D.D., Grattan, J.P. & Schwenniger, J.-L.(1996) The Quaternary geology of the coasts of theislands of the southern Outer Hebrides: a stratigraphicsurvey of the Holocene coastal dune and machairsequences. In: The Environment of the Outer Hebridesof Scotland: the last 14000 years (Eds D.D. Gilbertson,M. Kent & J.P. Grattan), pp. 59–71. Sheffield AcademicPress, Sheffield.

Gilfillan, E.S., Maher, N.P., Krejsa, C.M., et al.(1995) Use of remote sensing to document changes in marsh vegetation following the Amoco Cadiz oil spill(Brittany, France, 1978). Marine Pollution Bulletin30, 780–7.

Gill, A.B. (2005) Offshore renewable energy: ecologicalimplications of generating electricity in the coastalzone. Journal of Applied Ecology 42, 605–15.

ESD01 9/5/06 3:09 PM Page 400

REFERENCES 401

Gilvear, D.J. (1999) Fluvial geomorphology and riverengineering: future role utilizing a fluvial hydrosystemsframework. Geomorphology 31, 229–45.

Ginsburg, R.N. (1957) Early diagenesis and lithificationof shallow-water carbonate sediments in south Florida.In: Regional Aspects of Carbonate Deposition (EdsR.J. LeBlanc & J.G. Breeding), pp. 80–100. SpecialPublication 5, Society of Economic Paleontologistsand Mineralogists, Tulsa, OK.

Gischler, E. & Lomando, A.J. (1999) Recent sedimentaryfacies of isolated carbonate platforms, Belize-Yucatansystem, Central America. Journal of SedimentaryResearch 69, 747–63.

Gjessing, E.T. (1976) Physical and Chemical Character-istics of Aquatic Humus. Ann Arbor Science Publication,120 pp.

Glade, T. (2003) Landslide occurrence as a response to land use change: a review of evidence from NewZealand. Catena 51, 297–314.

Gladstone, W., Tawfiq, N., Nasr, D., et al. (1999)Sustainable use of renewable resources and conserva-tion in the Red Sea and Gulf of Aden: issues, needs andstrategic actions. Ocean & Coastal Management 42,671–97.

Glennie, K.W. (1987) Desert sedimentary environments,past and present – a summary. Sedimentary Geology50, 135–66.

Glynn, P.W. (1993) Monsoonal upwelling and episodicAcanthaster predation as probable controls of coralreef distribution and community structure in Oman,Indian Ocean. Atoll Research Bulletin 379, 1–66.

Glynn, P.W. (1996) Coral reef bleaching: facts, hypothesesand implications. Global Change Biology 2, 495–509.

Goldstone, M.E., Kirk, P.W.W. & Lester, J.N. (1990)The behaviour of heavy metals during wastewatertreatment. II. Lead, nickel and zinc. Science of theTotal Environment 95, 253–70.

Golterman, H.L., Sly, P.G. & Thomas, R.L. (1983)Study of the Relationship between Water QualitySand Sediment Transport. Technical. Paper No. 26,Unesco, Paris, 231 pp.

Golubic, S., Perkins, R.D. & Lukas, K.J. (1975) Boringmicroorganisms and microborings in carbonate sub-strates. In: The Study of Trace Fossils (Ed. R.W. Frey),pp. 229–59. Springer-Verlag, Berlin.

Goodbred, S.L.Jr. (2003) Response of the Ganges dis-persal system to climate change: a source-to-sink viewsince the last interstade. Sedimentary Geology 162,83–104.

Goodwin, T.H., Young, A.R., Holmes, M.G.R., et al.(2003) The temporal and spatial variability of sedi-ment transport and yields within the Bradford Beckcatchment, West Yorkshire, The Science of the TotalEnvironment 314–16, 475–94.

Gore, J.A. (1985) The Restoration of Rivers andStreams. Theories and Experience. Ann Arbor Science(Butterworth), Stoneham, MA.

Gottofrey, J. (1990) The disposition of cadmium, nickel,mercury and methylmercury in fish and effects oflipophilic metal chelation. Thesis, SLU, Uppsala.

Götz, A. & Zimmermann, M. (1993) The 1991 rockslides in Randa: causes and consequences. LandslideNews 7(3), 22–5.

Goudie, A. (Ed.) (1981) Geomorphological Techniques.Allen and Unwin, London, 395 pp.

Goudie, A.S. (1977) Sodium sulphate weathering andthe disintegration of Mehenjo-Daro, Pakistan. EarthSurface Processes 2, 75–86.

Goudie, A.S. (1984) Salt efflorescence and salt weather-ing in the Jurza Valley, Karakoram Mountains, Pakistan.In: The International Karakoram Project (2 Vols) (Ed.K.J. Miller), pp. 607–15. Cambridge University Press,Cambridge.

Goudie, A.S. (1989) Wind erosion in deserts. Proceed-ings of the Geologists Association 100, 89–92.

Goudie, A.S. (1997) Weathering processes. In: AridZone Geomorphology: Process, Form and Change inDrylands, 2nd edn (Ed. D.S.G. Thomas), pp. 25–39.Wiley, Chichester.

Goudie, A.S. & Wells, G.L. (1995) The nature, distribu-tion and formation of pans in arid zones. EarthScience Reviews 38, 1–69.

Goudie, A.S. & Wilkinson, J. (1978) The Warm DesertEnvironment. Cambridge University Press, Cambridge,88 pp.

Goy, J.L., Zazo, C. & Dabrio, C.J. (2003) A beach-ridgeprogradation complex reflecting periodical sea-leveland climate variability during the Holocene (Gulf ofAlmeria, Western Mediterranean). Geomorphology50, 251–68.

Graham, J. (1988) Collection and analysis of field data.In: Techniques in Sedimentology (Ed. M.E. Tucker),pp. 5–62. Blackwell Science, Oxford.

Grigg, R.W. & Dollar, S.J. (1990) Natural and anthro-pogenic disturbance on coral reefs. In: Ecosystems of the World 25, Coral Reefs (Ed. Z. Dubinsky), pp. 439–52. Elsevier, Amsterdam.

Grimalt, J.O., Ferrer, M. & Macpherson, E. (1999) The mine tailings accident in Aznalcollar. Science ofthe Total Environment 242, 3–11.

Grindrod, J. & Rhodes, E.G. (1984) Holocene sea-levelhistory of a tropical estuary: Missionary Bay, northQueensland. In: Coastal Geomorphology in Australia(Ed. B.G. Thom), pp. 151–78. Academic Press, Sydney.

Gromaire, M.C., Garnaud, S.S., Saard, M., et al. (2001)Contribution of different sources to the pollution ofwet weather flow in combined sewers. Water Research35, 521–33.

Gromaire-Mertz, M.C., Garnaud, S.S., Gonzalez, A., et al. (1999) Characteristics of urban runoff pollutionin Paris, Water Science and Technology 39, 1–8.

Grove, A.T. (1977) The geography of semi-arid lands.Philosophical Transactions of the Royal Society ofLondon, Series B, 278, 457–75.

ESD01 9/5/06 3:09 PM Page 401

402 REFERENCES

Guinotte, J., Buddemeier, R.W. & Kleypas, J. (2003)Future coral reef habitat marginality: temporal andspatial effects of climate change in the Pacific basin.Coral Reefs 22, 551–8.

Gupta, A. & Ahmad, R. (1999) Geomorphology and theurban tropics: building an interface between researchand usage. Geomorphology 31, 133–49.

Gustafsson, Ö. & Gschwend, P.M. (1997) Aquatic colloids: Concepts, definitions and current challenges.Limnology Oceanography 42, 519–28.

Guzmán, H.M., Burns, K.A. & Jackson, J.B.C. (1994)Injury, regeneration and growth of Caribbean reefcorals after a major oil spill in Panama. Marine EcologyProgress Series 105, 231–41.

Gygi, R.A. (1975) Sparisoma viride (Bonnaterre), theSpotlight Parrotfish, a major sediment producer oncoral reefs of Bermuda. Ecologiae Geologiae Helvetiae68, 327–59.

Haas, J.N., Richoz, I., Tinner, W., et al. (1998) Syn-chronous Holocene climatic oscillations recorded onthe Swiss Plateau and at timberline in the Alps. TheHolocene 8, 301–9.

Hadley Centre (2004) Hadley Centre for ClimatePrediction and Research. http://www.metoffice.com/research/hadleycentre/models/modeldata.html(accessed 29 July 2004).

Haeberli, W. (1995) Climate change impacts on glaciersand permafrost. In: Potential Ecological Impacts of Climate Change in the Alps and FennoscandianMountains (Eds A. Guisan, J.I. Holten, R. Spichiger & L. Tessiers), pp. 97–103. Conservatoire et JardinBotaniques, Geneva.

Haeberli, W., Rickenmann, D., Zimmermann, M., et al.(1990) Investigation of 1987 debris flows in the SwissAlps: general concept and geophysical soundings. In:Hydrology in Mountainous Regions. II – ArtificialReservoirs; Water and Slopes (Eds R.O. Sinniger &M. Monbaron), pp. 303–10. IAHS Publication 194,Proceedings of two Symposia held at Lausanne, August.International Association of Hydrological SciencesPress, Wallingford.

Haeberli, W., Kääb, A., Paul, F., et al. (2002) A surge-type movement at Ghiacciaio del belevedere and adeveloping slope instability in the east face of MonteRosa, Macugna, Italian Alps. Norwegian Journal ofGeography 56, 104–11.

Håkanson, L. (1977) The influence of wind, fetch, andwater depth on the distribution of sediments in LakeVänern, Sweden. Canadian Journal of Earth Science14, 397–412.

Håkanson, L. (1980) An ecological risk index for aquaticpollution control – a sedimentological approach. WaterResearch 14, 995–1001.

Håkanson, L. (1995) Models to predict organic contentof lake sediments. Ecological Modelling 82, 233–45.

Håkanson, L. (1999) Water Pollution – Methods andCriteria to Rank, Model and Remediate Chemical

Threats to Aquatic Ecosystems. Backhuys Publishers,Leiden, 299 pp.

Håkanson, L. (2004) Lakes – Form and Function. TheBlackburn Press, New Jersey.

Håkanson, L. & Boulion, V. (2002) The Lake Foodweb– Modelling Predation and Abiotic/Biotic Interactions.Backhuys Publishers, Leiden, 344 pp.

Håkanson, L. & Jansson, M. (1983) Principles of LakeSedimentology. Springer-Verlag, Berlin, 316 pp.

Håkanson, L. & Peters, R.H. (1995) Predictive Limno-logy. Methods for Predictive Modelling. SPB AcademicPublishing, Amsterdam, 464 pp.

Håkanson, L., Borg, H. & Uhrberg, R. (1990) Reliabil-ity of analyses of Hg, Fe, Ca, K, P, pH, alkalinity, conductivity, hardness and colour from lakes. Inter-national Reviews in Hydrobiology 75, 79–94.

Håkanson, L., Parparov, A. & Hambright, K.D. (2000)Modelling the impact of water level fluctuations onwater quality (suspended particulate matter) in LakeKinneret, Israel. Ecological Modelling 128, 101–25.

Halfar, J., Godinez-Orta, L. & Ingle, J.C. (2000) Micro-facies analysis of Recent carbonate environments inthe southern Gulf of California, Mexico – a model forwarm-temperate to subtropical carbonate formation.Palaios 15, 323–42.

Hamilton, R.S., Revitt, D.M., Warren, R.S. (1984)Levels and physico-chemical association of Cd, Cu, Pband Zn in road sediments. The Science of the TotalEnvironment 33, 59–74.

Hamilton, S.K. (1999) Potential effects of a major navigation project (Paraguay-Paraná Hidrovía) oninundation in the Pantanal floodplains. RegulatedRivers – Research and Management 15, 289–99.

Hammarlund, D., Björck, S., Buchardt, B., et al. (2003)Rapid hydrological changes during the Holocencerevealed by stable isotope records of lacustrine carbon-ates from Lake Igelsjön, southern Sweden. QuaternaryScience Reviews 22, 353–70.

Hammond, L.S. (1981) An analysis of grain size modifica-tion in biogenic carbonate sediments by deposit-feeding holothurians and echinoids (Echinodermata).Limnology and Oceanography 26, 898–906.

Hampel, R. (1980) Geschiebeberechnung für Gefah-renzonenpläne in Wildbachgebieten. Interpraevent 4,83–92.

Hansen, K. (1961) Lake types and lake sediments.Verhandlungen der Internationalen Vereinigung derLimnologie 14, 285–90.

Harbison, P. (1986) Mangrove muds – a sink and asource for trace metals. Marine Pollution Bulletin 17,246–50.

Harbor, J. & Warburton, J. (1993) Relative rates ofglacial and nonglacial erosion in Alpine environments.Arctic and Alpine Research 25(1), 1–7.

Hardisty, J. & Whitehouse, R.J.S. (1988) Evidence for a new sand transport process from experiments onSaharan dunes. Nature 332, 532–4.

ESD01 9/5/06 3:09 PM Page 402

REFERENCES 403

Hardman-Mountford, N.J., Allen, J.I., Frost, M.T., et al.(2005) Diagnostic monitoring of a changing environ-ment: an alternative UK perspective. Marine PollutionBulletin 50, 1463–71.

Harney, J.N. & Fletcher, C.H. (2003) A budget of car-bonate framework and sediment production, KailuaBay, Oahu, Hawaiian Islands. Journal of SedimentaryResearch 73, 856–68.

Harris, P.T. (1989) Sandwave movement under tidal andwind-driven currents in a shallow marine environment:Adolphus Channel, northeastern Australia. ContinentalShelf Research 9, 981–1002.

Harris, P.T. (1991) Reversal of subtidal dune asym-metries caused by seasonally reversing wind-drivencurrents in Torres Strait, northeastern Australia.Continental Shelf Research 11, 655–62.

Harris, P.T., Tsuji, Y., Marshall, J.F., et al. (1996) Sandand rhodolith-gravel entrainment on the mid- toouter-shelf under a western boundary current: FraserIsland continental shelf, eastern Australia. MarineGeology 129, 313–30.

Harrison, R.M. (1979) Toxic metals in street and house-hold dusts. The Science of the Total Environment 11,89–97.

Harrison, R.M., Yin, J. & Mark, D. (2001) Studies of the coarse particle (2.5–10 μm) component in UKurban atmospheres. Atmospheric Environment 35,3667–79.

Hartley, A.J. (2003) Andean uplift and climate change.Journal of the Geological Society 160, 7–10.

Hartley, A.J. & May, G. (1998) Miocene gypcretes fromthe Calama Basin, northern Chile. Sedimentology45, 351–64.

Hartley, A.J., Chong, G., Houston, J., et al. (2005) 150 million years of climatic stability: evidence fromthe Atacama Desert, northern Chile. Journal of theGeological Society, London 162, 421–4.

Harvey, A.M. (1984) Debris flow and fluvial deposits in Spanish Quaternary alluvial fans: implications for fan morphology. In: Sedimentology of Gravels and Conglomerates (Eds E.H. Koster & R. Steel), pp. 123–32. Memoir 10, Canadian Society of PetroleumGeologists, Calgary, Alberta.

Harvey, A.M. (1987) Patterns of Quaternary aggrada-tional and dissectional landform development in theAlmería region, southeast Spain: a dry region tectonic-ally active landscape. Die Erde 118, 193–215.

Harvey, A.M. (1990) Factors influencing Quaternaryalluvial fan development in southeast Spain. In:Alluvial Fans: a Field Approach (Eds A.H. Rachocki& M. Church), pp. 247–69. Wiley, Chichester.

Harvey, A.M. (1997) The occurrence and role of aridzone alluvial fans. In: Arid Zone Geomorphology:Process, Form and Change in Drylands, 2nd edn (Ed. D.S.G. Thomas), pp. 136–58. Wiley, Chichester.

Harvey, A.M., Foster, G., Hannam, J., et al. (2003) The Tabernas alluvial fan and lake system, southeast

Spain: applications of mineral magnetic and pedogeniciron oxide analyses towards clarifying the Quaternarysediment sequences. Geomorphology 50, 151–71.

Haslett, S.K. (2000) Coastal Systems. Routledge, London.Haslett, S.K., Cundy, A.B., Davies, C.F.C., et al. (2003)

Salt marsh sedimentation over the past c. 120 yearsalong the west Cotentin Coast of Normandy (France):Relationship to sea-level rise and sediment supply.Journal of Coastal Research 19(3), 609–20.

Hassan, M.A. (1990) Observations of desert flood bores.Earth Surface Processes and Landforms 15, 481–5.

Hassan, M.A. & Church, M. (1992) The movement ofindividual grains on the streambed. In: Dynamics ofGravel-bed Rivers (Eds P. Billi, R.D. Hey, C.R. Throne& P. Tacconi), pp. 159–73. Wiley, Chichester.

Hassan, M.A. & Klein, M. (2002) Fluvial adjustment ofthe Lower Jordan River to a drop in the Dead Sealevel. Geomorphology 45, 21–33.

Hatton, J.C. & Couto, A.L. (1992) The effect of coastlinechange on mangrove community structure, PortugueseIsland, Mozambique. Hydrobiologia 247, 49–57.

Hayes, M.O. (1979) Barrier island morphology as afunction of wave and tide regime. In: Barrier Islandsfrom the Gulf of St. Lawrence to the Gulf of Mexico(Ed. S.P. Leatherman), pp. 1–29. Academic Press,New York.

Hayes, M.O. (1991) Geomorphology and sedimentationpatterns in tidal inlets – a review. Coastal Sediments’91, Proceedings of a Special Conference on Quantita-tive Approaches to Coastal Sediment Processes, Vol. II(Eds N.C. Kraus, K.J. Gingerich & D.L. Kriebel). pp. 1343–55. American Society of Civil EngineersProceedings, Reston, VA.

Hayes, S.K., Montgomery, D.R. & Newhall, C.G. (2002)Fluvial sediment transport and deposition followingthe 1991 eruption of Mount Pinatubo. Geomorphology45, 211–24.

Hayne, M. & Chappell, J. (2001) Cyclone frequencyduring the last 5000 years at Curacoa Island, northQueensland, Australia. Palaeogeography, Palaeo-climatology, Palaeoecology 168, 207–19.

Haynes, D. & Johnson, J.E. (2000) Organochlorine,heavy metal and polyaromatic hydrocarbon pollutantconcentrations in the Great Barrier Reef (Australia)environment: a review. Marine Pollution Bulletin 41, 267–78.

Haynes, F.N. & Coulson M.G. (1982) The decline of Spartina in Langstone Harbour, Hampshire.Hampshire Field Club and Archaeological Society38, 5–18.

Hearn, G.J. (1995) Landslide and erosion hazard map-ping at Ok-Tedi copper mine, Papua-New-Guinea.Quaterly Journal of Engineering Geology 28, 47–60.

Heathcote, R.L. (1983) The Arid Lands: their Use andAbuse. Longman, London, 323 pp.

Hegge, B.J., Eliot, I.G. & Hsu, J. (1996) Sheltered sandybeaches. Journal of Coastal Research 12, 748–60.

ESD01 9/5/06 3:09 PM Page 403

404 REFERENCES

Heim, A. (1882) Der Bergsturz von Elm. Deutsche Geo-logische Gesellschaft für Zeitschrift 34, 74–115.

Helgen, S.O. & Moore, J.N. (1996) Natural backgrounddetermination and impact quantification in trace metal-contaminated river systems. Environmental Scienceand Technology 30, 129–35.

Hellden, U. (1988) Desertification monitoring: is thedesert encroaching? Desertification Control Bulletin17, 8–12.

Hem, J.E. (1972) Chemistry and occurrence of cadmiumand zinc in surface water and groundwater. WaterResearch 8, 661–79.

Hettler, J., Irion, G. & Lehmann, B. (1997) Environ-mental impact of mining waste disposal on a tropicallowland river system: a case study on the Ok Tedimine, Papua New Guinea. Mineralium Deposita 32,280–91.

Hewitt, K. (2004) Geomorphic Hazards in MountainEnvironments. In: Mountain Geomorphology (EdsP.N. Owens & O. Slaymaker), pp. 187–218. Arnold,London.

Hindson, R.A. & Andrade, C. (1999) Sedimentation and hydrodynamic processes associated with thetsunami generated by the 1755 Lisbon earthquake.Quaternary International 56, 27–38.

Hinrichsen, D. (1998) Coastal Waters of the World:Trends, Threats, and Strategies. Island Press,Washington, DC, 420 pp.

Hjalmarson, H.W. (1984) Flash flood in Tanque Verde Creek, Tucson, Arizona. Journal of HydraulicEngineering, American Society of Civil Engineers110, 1841–52.

Hjulström, F. (1935) Studies on the morphological activity of rivers as illustrated by the River Fyris.Bulletin of the Geological Institute of Uppsala 25,221–527.

Hogarth, P.J. (1999) The Biology of Mangroves. OxfordUniversity Press, Oxford, pp. 228.

Holme, N.A. (1961) The bottom fauna of the EnglishChannel. Journal of the Marine Biological Associationof the United Kingdom 41, 397–461.

Holme, N.A. (1966) The bottom fauna of the EnglishChannel. Part II. Journal of the Marine BiologicalAssociation of the United Kingdom 46, 401–93.

Holme, N.A. & Wilson, J.B. (1985) Faunas associ-ated with longitudinal furrows and sand ribbons in atide-swept area in the English Channel. Journal of theMarine Biological Association of the United Kingdom65, 1051–72.

Hooke, R.L. (2000) Towards a uniform theory of clasticsediment yield in fluvial systems. Geological Society ofAmerica Bulletin 112, 1778–86.

Hopke, P.K., Lamb, R.E. & Natusch, F.S. (1980) Multi-elemental characterization of urban roadway dust.Environmental Science and Technology 14, 164–72.

Horowitz, A.J. (1991) A Primer on Sediment Trace-element Chemistry. Lewis Publishers, Michigan.

Horowitz, A.J. & Elrick, K.A. (1987) The relation ofstream sediment surface area, grain size and composi-tion to trace element chemistry. Applied Geochemistry2, 437–51.

Hosiokangas, J., Ruuskanen, J. & Pekkanen, J. (1999)Effects of soil dust episodes and mixed fuel sources onsource apportionment of PM10 particles in Kuopio,Finland. Atmospheric Environment 33, 3821–9.

Hossain, S. & Eyre, B. (2002) Suspended sedimentexchange through the sub-tropical Richmond RiverEstuary, Australia: a balance approach. Estuarine,Coastal and Shelf Science 55, 579–86.

Houghton, J.T., Callander, B.A. & Varney, S.K. (1992) Climate Change. Cambridge University Press,Cambridge, pp. 69–95.

Hovland, M. & Judd, A.G. (1988) Seabed Pockmarksand Seepages. Graham and Trotman, London.

Hsu, K.H. (1975) Catastrophic debris streams(sturzstroms) generated by rockfalls. Geological Societyof America Bulletin 86, 129–40.

Hubbard, D.K. (1992) Hurricane-induced sedimenttransport in open-shelf tropical systems – an examplefrom St. Croix, U.S. Virgin Islands. Journal ofSedimentary Petrology 62, 946–60.

Hubbard, D.K. (1997) Reefs as dynamic systems. In:Life and Death of Coral Reefs (Ed. C. Birkeland), pp. 43–67. Chapman and Hall, London.

Hudson-Edwards, K.A., Macklin, M.G., Curtis, C.D., et al. (1996) Processes of formation and distribu-tion of Pb-, Zn-, Cd- and Cu-bearing minerals in theTyne basin, NE England: implications for metal-contaminated river systems. Environmental Scienceand Technology 30, 72–80.

Hudson-Edwards, K.A., Macklin, M.G., Curtis, C.D., et al. (1998) Chemical remobilisation of contaminantmetals within floodplain sediments in an incising river system: implications for dating and chemostrati-graphy. Earth Surface Processes and Landforms 23,671–84.

Hudson-Edwards, K.A., Macklin, M.G. & Taylor, M.P.(1999a) 2000 years of sediment-borne heavy metalstorage in the Yorkshire Ouse basin, NE England.Hydrological Processes 13, 1087–102.

Hudson-Edwards, K.A., Schell, C. & Macklin, M.G.(1999b) Mineralogy and geochemistry of alluviumcontaminated by metal mining in the Rio Tinto area, southwest Spain. Applied Geochemistry 14,55–70.

Hudson-Edwards, K.A., Macklin, M.G., Miller, J.R. &Lechler, P.J. (2001) Sources, distribution and storageof heavy metals in the Río Pilcomayo, Bolivia. Journalof Geochemical Exploration 72, 229–50.

Hudson-Edwards, K.A., Macklin, M.G., Jamieson, H.E.,et al. (2003) The impact of tailings dam spills andclean-up operations on sediment and water quality inriver systems: the Ríos Agrio-Guadiamar, Aznalcóllar,Spain. Applied Geochemistry 18, 221–39.

ESD01 9/5/06 3:09 PM Page 404

REFERENCES 405

Hulme, M. (1992) Rainfall changes in Africa: 1931–60to 1961–90. International Journal of Climatology12, 685–99.

Huston, M. (1985) Variation in coral growth rates with depth at Discovery Bay, Jamaica. Coral Reefs4, 19–25.

Hutchings, P.A. (1986) Biological destruction of coralreefs. Coral Reefs 4, 239–52.

Hutchinson, G.E. (1957) A Treatise on Limnology,Vol. 1. Geography, Physics and Chemistry. Wiley,New York, 1016 pp.

Hutchinson, G.E. (1967) A Treatise on Limnology. Vol. 2. Introduction to Lake Biology and the Limno-plankton. Wiley, New York, 1115 pp.

Hutchinson, J.N. (1988) General Report: Morphologicaland geotechnical parameters of landslides in relationto geology and hydrogeology. In: Proceedings of the5th International Symposium on Landslides, Vol. 1(Ed. C. Bonnard), pp. 3–35. Balkema, Rotterdam.

Hvitved-Jacobson, T., Yousef, Y.A., Wanielista, M.P., et al. (1984) Fate of phosphorus and nitrogen in pondsreceiving highway runoff. The Science of the TotalEnvironment 33, 259–70.

Inbar, M., Tamir, M. & Wittenberg, L. (1998) Runoffand erosion processes after a forest fire in MountCarmel, a Mediterranean area. Geomorphology 24,17–33.

Ingersoll, C.G., Haverland, P.S., Brunson, E.L., et al.(1996) Calculation of sediment effect concentrations.Journal of Great Lakes Research 22, 602–23.

Inman, D.L. & Jenkins, S.A. (1999) Climate change andthe episodicity of sediment flux of small Californianrivers. Journal of Geology 107, 251–70.

Inouchi, Y., Kinugasa, Y., Kumon, F., et al. (1996)Turbidites as records of intense paleoearthquakes in Lake Biwa, Japan. Sedimentary Geology 104,117–25.

Ives, J.D. & Messerli, B. (1989) The Himalayan Dilemma– Reconciling Development and Conservation.Routledge, London.

Jäckli, H. (1957) Gegenwartsgeologie des undnerischenRheingebietes-ein Beitrag zur exogenen Dynamikalpiner Gebirgslandschaften. Beitraege zur Geologieder Schweiz, Geotechnische Serie 36.

James, A. (1999) Time and persistence of alluvium:River engineering, fluvial geomorphology, and miningsediment in California. Geomorphology 31, 265–90.

James, C.S. (1985) Sediment transfer to overbank sections.Journal of Hydraulic Research 23, 435–52.

James, L.A. (1989) Sustained storage and transport ofhydraulic gold mining sediment in the Bear River,California. Annals of the Association of AmericanGeographers 79, 570–92.

James, N.P. & Bourque, P.A. (1992) Reefs and mounds.In: Facies Models: Response to Sea Level Change (EdsR.G. Walker & N.P. James), pp. 323–47. GeologicalAssociation of Canada, St John’s, Newfoundland.

James, N.P. & Macintyre, I.G. (1985) CarbonateDepositional Environments Modern and Ancient.Colorado School of Mines Quarterly 80.

James, N.P., Bone, Y., Von de Borsch, C.C., et al.(1992) Modern carbonate and terrigenous clastic sediments on a cool water, high energy, mid-latitudeshelf: Lacepede, southern Australia. Sedimentology38, 877–903.

Janda, R.J., Daag, A.S., Delos Reyes, P.J., et al. (1996)Assessment and response to lahar hazard aroundMount Pinatubo, 1991 to 1993. In: Fire and Mud,Eruptions and Lahars of Mount Pinatubo, Philippines(Eds C.G. Newhall & R.S. Punongbayan), pp. 107–39.PHIVOLCS Press, Quezon City, and University ofWashington Press, Seattle.

Jansen, J.M.L. & Painter, R.B. (1974) Predicting sedi-ment yield from climate and topography. Journal ofHydrology 21, 371–80.

Jansson, M.B. (1988) A global survey of sediment yield.Geografiska Annaler 70A(1–2), 81–98.

Jennings, S. & Kaiser, M.J. (1998) The effects of fishingon marine ecosystems. Advances in Marine Biology34, 203–352.

Johnson, D.P. & Risk, M.J. (1987) Fringing reef growthon a terrigenous mud foundation, Fantome Island,central Great Barrier reef, Australia. Sedimentology34, 275–87.

Johnson, H.D. & Baldwin, C.T. (1996) Shallow clastic seas. In: Sedimentary Environments: Processes, Facies and Stratigraphy, 3rd edn (Ed. H.G. Reading),pp. 232–80. Blackwell Science, Oxford.

Johnsson, M.J., Stallard, R.F. & Lundberg, N. (1991) Controls on the composition of fluvial sandsfrom a tropical weathering environment – sands of the Orinoco River drainage-basin, Venezuela andColumbia. Geological Society of America Bulletin103, 1622–47.

Jones, B.F. & Bowser, C.J. (1978) The mineralogy andrelated chemistry of lake sediments. In: Lakes: Chem-istry, Geology, Physics (Ed. A. Lerman), pp. 179–235.Springer-Verlag, Berlin.

Jones, D.K.C. (1992) Landslide hazard assessment in the context of development. In: Geohazards: Naturaland Man-made (Eds G.J.H. McCall, D.J.C. Laming & S.C. Scott), pp. 117–41. Chapman and Hall,London.

Jones, D.K.C. (1993a) Global warming and geomorpho-logy. The Geographical Journal 159, 124–30.

Jones, D.K.C. (1993b) Slope instability in a warmerBritain. The Geographical Journal 159, 184–95.

Jones, D.K.C., Brunsden, D. & Goudie, A.S. (1983) A preliminary geomorphological assessment of theKarakoram Highway. Quarterly Journal of Engineer-ing Geology 16, 331–55.

Jonsson, P. (1992) Large-scale changes of contaminantsin Baltic Sea sediments during the twentieth century.Thesis, Uppsala University, Sweden.

ESD01 9/5/06 3:09 PM Page 405

406 REFERENCES

Judd, A., Franklin, F. & Faire, S. (Eds) (2003) OSPARWorkshop – Environmental Assessment of RenewableEnergy in the Marine Environment, Maldon, 17–18September, 109 pp.

Kääb, A. (2002) Monitoring high-mountain terraindeformation from repeated air and spaceborne opticaldata: examples using digital aerial imagery and ASTERdata. ISPRS Journal of Photogrammetry and RemoteSensing 57, 39–52.

Kääb, A., Wessels, R., Haeberli, W., et al. (2003) Rapid ASTER imaging facilitates timely assessment ofglacier hazards and disasters. Eos (Transactions of theAmerican Geophysical Union) 84, 13,117.

Kaiser, M.J. & De Groot, S.J. (2000) The Effects ofFishing on Non-target Species and Habitats: Biological,Conservation and Socio-economic issues. BlackwellScience, Oxford.

Karrickhoff, S.W. (1981) Semi-empirical estimation ofsorption of hydrophobic pollutants on natural sedi-ments and soils. Chemosphere 10, 833–46.

Karwan, D.L., Allan, J.D. & Bergen, K.M. (2001) Chang-ing near-stream land use and river channel morpho-logy in the Venezuelan Andes. Journal of the AmericanWater Resources Association 37, 1579–87.

Kasse, C. (1998) Depositional model for cold-climatetundra rivers. In: Palaeohydrology and EnvironmentalChange (Eds G. Benito, V.R. Baker & K.J. Gregory),pp. 83–97. Wiley, Chichester.

Kasse, C., Vandenberghe, J., Van Huissteden, J., Bohncke,S.J.P., et al. (2003) Sensitivity of Weichselian fluvialsystems to climate change (Nochten mine, easternGermany). Quaternary Science Reviews 22, 2141–56.

Kato, M., Fukusawa, H. & Yasuda, Y. (2003) Varvedlacustrine sediments of Lake Tougou-ike, WesternJapan, with reference to Holocene sea-level changes in Japan. Quarternary International 105, 33–7.

Ke, L., Wang, W.Q., Wong, T.W.Y., et al. (2003)Removal of pyrene from contaminated sediments bymangrove microcosms. Chemosphere 51, 25–34.

Kearney, M.S. & Stevenson, J.C. (1991) Island land lossand marsh vertical accretion rate evidence for histor-ical sea-level changes in Chesapeake Bay. Journal ofCoastal Research 7, 403–15.

Keen, T.R. & Slingerland, R.L. (1993) Four storm-eventbeds and the tropical cyclones that produced them: anumerical hindcast. Journal of Sedimentary Petrology63, 218–32.

Kelderman, P., Drossaert, W.M.E., Min, Z., et al. (2000)Pollution assessment of the canal sediments in the city of Delft (the Netherlands). Water Research 34,936–44.

Kemp, A.L.W., Thomas, R.L., Dell, C.I., et al. (1976)Cultural impact of the geochemistry of sediments inLake Erie. Journal of the Fisheries Research Board ofCanada (Special Issue) 33, 440–62.

Kench, P.S. (1997) Physical processes in an Indian Oceanatoll. Coral Reefs 17, 155–68.

Kench, P.S. & Cowell, P.J. (2002) Variations in sedi-ment production and implications for atoll island stability under rising sea level. Proceedings of 9thinternational Coral Reef Symposium, Bali, Vol. 2,pp. 1181–6. Ministry of Environment (Indonesia),Indonesian Institute of Sciences and InternationalSociety for Reef Studies.

Kench, P.S. & McLean, R.F. (1996) Hydraulic charac-teristics of bioclastic deposits: new possibilities forenvironmental interpretation using settling velocityfractions. Sedimentology 43, 561–70.

Kennedy, D.M. & Woodroffe, C.D. (2002) Fringing reef growth and morphology: a review. Earth-ScienceReviews 57, 255–77.

Kenny, A.J., Rees, H.L., Greening, J., et al. (1998) Theeffects of gravel extraction on the macrobenthos at an experimental dredge site off North Norfolk, UK(results 3 years post-dredging). International Councilfor the Exploration of the Seas, Council Minutes14, 1–7.

Kenny, A.J. & Rees, H.L. (1994) The effects of marinegravel extraction on the macrobenthos: Early post-dredging recolonization. Marine Pollution Bulletin28, 442–7.

Kenny, A.J. & Rees, H.L. (1996) The effects of marinegravel extraction on the macrobenthos: results 2 yearspost-dredging. Marine Pollution Bulletin 32, 615–22.

Keown, M.P., Dardeau, E.A. & Causey, E.M. (1986)Historic trends in the sediment flow regime of theMississippi River. Water Resources Research 22,1555–64.

Kershaw, P.J., Swift, D.J. & Denoon, D.C. (1988)Evidence of recent sedimentation in the eastern IrishSea. Marine Geology 85, 1–14.

Kim, K.W., Myung, J.H., Ahn, J.S., et al. (1998) Heavymetal contamination in dusts and stream sediments in the Taejon Area, Korea. Journal of GeochemicalExploration 64, 409–19.

Kineke, G.C., Sternberg, R.W., Trowbridge, J.H., et al. (1996) Fluid-mud processes on the Amazon continental shelf. Continental Shelf Research 16,667–96.

Kirchner, J.W., Finkel, R.C., Riebe, C.S., et al. (2001)Mountain erosion over 10 yr, 10 k.y., and 10 m.y.time scales. Geology 29, 591–4.

Kirkby, M.J., Abrahart, R., McMahon, M.D., Shao, J. &Thornes, J.B. (1998) MEDALUS soil erosion modelsfor global change. Geomorphology 24, 35–49.

Kleypas, J. & Langdon, C. (2002) Overview of CO2-induced changes in seawater chemistry. Proceedingsof 9th International Coral Reef Symposium, Bali, Vol.2, pp. 1085–9. Ministry of Environment (Indonesia),Indonesian Institute of Sciences and InternationalSociety for Reef Studies.

Kleypas, J.A., McManus, J.W. & Meñez, L.A.B. (1999)Environmental limits to coral reef development: wheredo we draw the line? American Zoologist 39, 146–59.

ESD01 9/5/06 3:09 PM Page 406

REFERENCES 407

Knighton, D. (1998) Fluvial Forms and Processes.Arnold, London.

Knighton, D. & Nanson, G. (1997) Distinctiveness,diversity and uniqueness in arid zone river systems. In: Arid Zone Geomorphology: Process, Form andChange in Drylands, 2nd edn (Ed. D.S.G. Thomas),pp. 185–203. Wiley, Chichester.

Knox, J.C. (1987) Historical valley floor sedimentationin the upper Mississippi valley. Annals of the Associ-ation of American Geographers 77, 224–44.

Knox, J.C. (1995) Fluvial systems since 20,000 years BP. In: Global Continental Palaeohydrology (Eds K.J. Gregory, L. Starkel & V.R. Baker), pp. 87–108.Wiley, Chichester.

Knox, J.C. (2001) Agricultural influence on landscapesensitivity in the Upper Mississippi River Valley. Catena42, 193–224.

Knudson, T. (1991) Mountain lake turns to desert.Sacramento Bee, 10 June.

Komar, P.D. (1998) Beach and Nearshore Sedimenta-tion, 2nd edn. Prentice Hall, Upper Saddle River, NJ.

Komatsubara, J. (2004) Fluvial architecture and sequencestratigraphy of the Eocene to Oligocene Iwaki Forma-tion, northeast Japan: channel-fills related to the sea-level change. Sedimentary Geology 168, 109–23.

Kondolf, G.M. (1995) Managing bedload sediment in regulated rivers: examples from California, USA.In: Natural and Anthropogenic Influences in FluvialGeomorphology (Eds J.E. Costa, A.J. Miller, K.W.Potter & P.R. Wilcock), pp. 125–213. Monograph 89,Geophysical Union of America, Washington, DC.

Kornman, B.A. & De Deckere, E.M.G.T. (1998)Temporal variation in sediment erodability and sus-pended sediment dynamics in the Dollard estuary. In: Sedimentary Processes in the Intertidal Zone (EdsK.S. Black, D.H. Paterson & A. Cramp), pp. 231–41.Special Publication 139, Geological Society PublishingHouse, Bath.

Kranck, K. (1973) Flocculation of suspended sedimentin the sea. Nature 24, 348–50.

Krank, K. (1975) Sediment deposition from flocculatedsuspensions. Sedimentology 22, 111–23.

Kranck, K. (1979) Particle matter grain-size character-istics and flocculation in a partially mixed estuary.Sedimentology 28, 107–14.

Kratzer, C.R. (1999) Transport of sediment-boundorganochlorine pesticides to the San Joaquin River,California. Journal of the American Water ResourcesAssociation 35, 957–81.

Krauss, K.W., Allen, J.A. & Cahoon, D.R. (2003)Differential rates of vertical accretion and elevationchange among aerial root types in Micronesian man-grove forests. Estuarine, Coastal and Shelf Science56, 251–9.

Krumbein, W.C. (1963) A geological process–responsemodel for analysis of beach phenomena. Bulletin ofthe Beach Erosion Board 17, 1–15.

Kuehl, S., DeMaster, D.J. & Nittrouer, C.A. (1986)Nature of sediment accumulation on the Amazon con-tinental shelf. Continental Shelf Research 6, 209–25.

Laban, P. (1978) Field Measurements on Erosion andSedimentation in Nepal. PAO/UNDP/IWM?SP/05,Department of Soil Conservation and WatershedManagement, Nepal.

Laban, P. (1979) Landslide Occurrence in Nepal. PhewaTal Project Report no. 8P/13, Integrated WatershedManagement Project, Kathmandu.

Lacerda, L.D. & Salomons, W. (1992) Mercurio naAmazonia: uma Bomba Relogio Quimica? Serie Tecno-logia Ambiental 3, CETEM/CNPq, Rio de Janeiro.

Laird, J.R. & Harvey, M.D. (1986) Complex response ofa chaparral drainage basin to fire. In: Drainage BasinSediment Delivery (Ed. R.F. Hadley), pp. 165–83.Publication 159, International Association of Hydro-logical Sciences, Washington, DC.

Lambeck, A. & Woolfe, K.J. (2000) Composition andtextural variability along the 10-metre isobath, GreatBarrier Reef – evidence for pervasive northward sedi-ment transport. Australian Journal of Earth Sciences47, 327–35.

Lamprey, H.F. (1975) Report on the Desert Encroach-ment Reconnaissance in Northern Sudan, October 21–November 10, 1975. National Council for Research,Ministry of Agriculture, Food and Resources,Khartoum, 16 pp.

Lancaster, N. (1994) Dune morphology and dynamics.In: Geomorphology of Desert Environments (EdsA.D. Abrahams & A.J. Parsons), pp. 474–505.Chapman and Hall, London.

Lancaster, N. (1995) The Geomorphology of DesertDunes. Routledge, London, 211 pp.

Land, L.S. (1970) Carbonate mud: production byepibiont growth on Thalassia testudinium. Journal ofSedimentary Petrology 42, 179–82.

Land, L.S. (1974) Growth rate of a West Indian(Jamaican) reef. Proceedings of Second InternationalCoral Reef Symposium, Vol. 2, pp. 409–12. TheGreat Barrier Reef Committee, Brisbane.

Langbein, W.B. & Schumm, S.A. (1958) Yield of sedimentin relation to mean annual precipitation. Transactionsof the American Geophysical Union 39, 1076–84.

Langedal, M. (1997a) Dispersion of tailings in theKnabeåna–Kvina drainage basin, Norway, 1: evalu-tion of overbank sediments as sampling medium forregional geochemical mapping. Journal of GeochemicalExploration 58, 157–72.

Langedal, M. (1997b) Dispersion of tailings in theKnabeåna–Kvina drainage basin, Norway, 2: mobilityof Cu and Mo in tailings-derived fluvial sediments.Journal of Geochemical Exploration 58, 173–83.

Larcombe, P. & Carter, R.M. (2004) Cyclone pump-ing, sediment partitioning and the development of theGreat Barrier Reef shelf system: a review. QuaternaryScience Reviews 23, 107–35.

ESD01 9/5/06 3:09 PM Page 407

408 REFERENCES

Larcombe, P. & Ridd, P.V. (1995) Megaripple dynamicsand sediment transport in a mesotidal mangrovecreek: implications for palaeoflow reconstructions.Sedimentology 42, 593–606.

Larcombe, P. & Woolfe, K.J. (1999) Terrigenous sedi-ments as influences upon Holocene nearshore coralreefs, central Great Barrier Reef, Australia. AustralianJournal of Earth Sciences 46, 141–54.

Larcombe, P., Ridd, P.V., Prytz, A. & Wilson, B. (1995)Factors controlling suspended sediment on inner-shelf coral reefs, Townsville, Australia. Coral Reefs14, 163–71.

Larcombe, P., Parker, R., Dye, S., et al. (2005) Marineenvironmental change over decades to millennia –evaluating proxy parameters and their potential use in understanding the state of the marine ecosystem.Final report to Defra.

Large, D.J., Fortey, N.J., Milodowski, A.E., et al. (2002)Petrographic observations of iron, copper and zincsulfides in freshwater canal sediment. Journal ofSedimentary Research 71, 61–9.

Laronne, J.B. & Reid, I. (1993) Very high bedload sediment transport in desert ephemeral rivers. Nature366, 148–50.

Laronne, J.B., Reid, I., Yitshak, Y., et al. (1994) Thenon-layering of gravel streambeds under ephemeralflow regimes. Journal of Hydrology 159, 353–63.

Lau, Y.L., Oliver, B.G. & Kirshnappan, B.G. (1989)Transport of some chlorinated contaminants by thewater, suspended sediments, and bed sediments in theSt. Clair and Detroit Rivers. Environmental Toxico-logy and Chemistry 8, 291–301.

Lave, J. & Avouac, J.P. (2001) Fluvial incision and tectonic uplift across the Himalayas of central Nepal.Journal of Geophysical Research – Solid Earth 106,26,561–91.

Lavee, H. & Poesen, J.W.A. (1991) Overland flow generation and continuity on stone-covered soil surfaces. Hydrological Processes 5, 345–60.

Lavier, L.L., Steckler, M.S. & Brigaud, F. (2001)Climatic and tectonic control on the Cenozoic evolu-tion of the West African margin. Marine Geology 178,63–80.

Le Houérou, H.N., Bingham, R.L. & Skerbek, W. (1988)Relationship between the variability of primary pro-duction and the variability of annual precipitation inworld arid lands. Journal of Arid Environments 15,1–18.

Lecce, S.A. & Pavlowsky, R.T. (2001) Use of mining-contaminated sediment tracers to investigate the timingand rates of historical flood plain sedimentation.Geomorphology 38, 85–108.

Lecoanet, H., Leveque, F. & Ambrosi, J.P. (2003)Combination of magnetic parameters: an efficient way to discriminate soil-contamination sources (southFrance). Environmental Pollution 122, 229–34.

Lee, S.K., Tan, W.H. & Havanond, S. (1996) Regenerationand colonisation of mangrove on clay-filled reclaimedland in Singapore. Hydrobiologia 319, 23–35.

Leeder, M. (1999) Sedimentology and SedimentaryBasins: from Turbulence to Tectonics. BlackwellScience, Oxford.

Leeder, M.R. (1999) Sedimentology and SedimentaryBasins. Blackwell Science, Oxford.

Leeks, G.J.L. & Roberts, G. (1987) The effects offorestry on upland streams – with special reference to water quality and sediment transport. In: Environ-mental Aspects of Plantation Forestry in Wales(Eds J.E.G. Good & Institute of Terrestrial Ecology), pp. 9–24. Publication Number 22, Institute ofTerrestrial Ecology, Huntingdon.

Lees, A. (1975) Possible influence of salinity and tem-perature on modern shelf carbonate sedimentation.Marine Geology 19, 159–98.

Legesse, D., Gasse, F., Radakovitch, O., et al. (2002)Environmental changes in a trophical lake (LakeAbiyata, Ethiopia) during recent centuries. Palaeo-geography, Palaeoclimatology, Palaeoecology 187,233–58.

Legleiter, C.J., Lawrence, R.L., Fonstad, M.A., et al.(2003) Fluvial response a decade after wildfire in thenorthern Yellowstone ecosystem: a spatially explicitanalysis. Geomorphology 54, 119–36.

Leigh, D.S. (1997) Mercury-tainted overbank sedi-ment from past gold mining in North Georgia, USA.Environmental Geology 30, 244–51.

Leonard, L., Clayton, T., Dixon, K., et al. (1989) Ananalysis of replenished beach design parameters onU.S. East Coast barrier islands: Journal of CoastalResearch 6, 15–36.

Leopold, L.B., Emmett, W.W. & Myrick, R.M. (1966)Channel and Hillslope Processes in a Semi-arid Areaof New Mexico. Professional Paper 352-G, USGeological Survey, Denver, CO.

Levings, S.C. & Garrity, S.D. (1994) Effects of oil spillson fringing red mangroves (Rhizophora mangle):losses of mobile species associated with submergedprop roots. Bulletin of Marine Science 54, 782–94.

Lewin, J. & Macklin, M.G. (1987) Metal mining andfloodplain sedimentation in Britain. In: InternationalGeomorphology 1986, Part 1 (Ed. V. Gardiner), pp.1009–27. Wiley, Chichester.

Lewin, J. & Wolfendon, P.J. (1978) The assessment ofsediment sources: a field experiment. Earth SurfaceProcesses 3, 171–8.

Lewin, J., Davies, B.E. & Wolfenden, P.J. (1977)Interactions between channel change and historic mining sediments. In: River Channel Changes (Ed.R.C. Gregory), pp. 353–67. Wiley, New York.

Lewin, J., Bradley, S.B. & Macklin, M.G. (1983)Historical valley alluviation in mid-Wales. GeologicalJournal 18, 331–50.

ESD01 9/5/06 3:09 PM Page 408

REFERENCES 409

Lewis, J.B. (2002) Evidence from aerial photography ofstructural loss of coral reefs at Barbados, West Indies.Coral Reefs 21, 49–56.

Lewis, R.R. (1983) Impacts of oil spills on mangroveforests. In: Tasks for Vegetation Science, Vol. 8 (Ed. H.J. Teas), pp. 171–83. Dr W. Junk Publishers,The Hague.

Lewis, S.G., Maddy, D. & Scaife, R.G. (2001) The fluvialsystem response to abrupt climate change during thelast cold stage: the Upper Pleistocene River Thamesfluvial succession at Ashton Keynes, UK. Global andPlanetary Change 28, 341–59.

Lick, W., Lick, J. & Ziegler, C.K. (1992) Flocculationand its effect on the vertical transport of fine-grainedsediments. Hydrobiologia 235/236, 1–16.

Liddell, W.D. & Ohlhorst, S.L. (1988) Hard substratacommunity patterns, 1–120 m, north Jamaica. Palaios3, 413–23.

Liddell, W.D. & Ohlhorst, S.L. (1993) Ten years of disturbance and change on a Jamaican fringing reef. Proceedings of Seventh International Coral ReefSymposium, Vol. 1, pp. 149–55. University of GuamPress, UOG Station, Guam.

Likens, G., Wright, R., Galloway, J., et al. (1979) Acidrain. Scientific American 241, 43–51.

Linton, R.W., Natusch, D.F.S., Solomon, R.L., et al.(1980) Physicochemical characterisation of lead inurban dusts. A microanalytical approach to lead tracing. Environmental Science and Technology 4,159–64.

List, J.H., Sallenger, A.H., Hansen, M.E., et al. (1997)Accelerated relative sea-level rise and rapid coastalerosion: testing a causal relationship for the Louisianabarrier islands. Marine Geology 140, 347–65.

Livingstone, I. & Warren, A. (1996) Aeolian Geomor-phology: an Introduction. Longman, London.

London Convention (2005) Compendium of MonitoringActivities at Disposal at Sea Sites in 2003. Submitted byCanada. London Convention, Scientific Group, 28thMeeting, 23–27 May 2005, paper LC/SG 28/INF.4.30 pp.

Long, E.D. & MacDonald, D.D. (1998) Recommendeduses of empirically derived, sediment quality guide-lines for marine and estuarine ecosystems. Human andEcological Risk Assessment 4, 1019–93.

Long, E.R., MacDonald, D.D., Smith, S.L., et al. (1995)Incidence of adverse biological effects within ranges of chemical concentrations in marine and estuarinesediments. Environmental Management 19, 81–97.

Longfield, S.A. & Macklin, M.G. (1999) The influenceof recent environmental change on flooding and sedi-ment fluxes in the Yorkshire Ouse basin. HydrologicalProcesses 13, 1051–66.

Longman, M.W. (1981) A process approach to recog-nizing facies of reef complexes. In: European FossilReef Models (Ed. D.F. Toomey), pp. 9–40. Special

Publication 30, Society for Economic Paleontologistsand Mineralogists, Tulsa, OK.

Louis, H. (1975) Neugefasstes Höhendiagramm der Erde,pp. 305–26. Bayerische Akademie der Wissensch-aften (Mathematisch-Naturwissenschaftliche KlasseSitzungsberichte), Munich.

Lovley, D.R. & Anderson, R.T. (2000) Influence of dissimilatory metal reduction on fate of organic andmetal contaminants in the subsurface. HydrogeologyJournal 8, 77–88.

Lovley, D.R. (1993) Dissimilatory metal reduction.Annual Reviews in Microbiology 47, 263–90.

Lucchitta, I. (1975) Application of ERTS Images andImage Processing to Regional Geologic Problems andGeologic Mapping in Northern Arizona: Part IVB, the Shivwitts Plateau. Technical Report 32-1597,National Aeronautical and Space Administration,Washington, DC, 41–72.

Ludlam, S.D. (1981) Sedimentation rates in FayettevilleGreen Lake, New York, U.S.A. Sedimentology 28,85–96.

Ly, C.K. (1980) The role of the Akosombo Dam on the Volta River in causing coastal erosion in centraland eastern Ghana (west Africa). Marine Geology37, 323–32.

Mabutt, J.A. (1978) The impact of desertification asrevealed by mapping. Environmental Conservation 5, 45–56.

Mabutt, J.A. (1985) Desertification of the world’s rangelands. Desertification Control Bulletin (UnitedNations Environmet Programme, Nairobi) 12, 1–11.

Mabutt, J.A. (1986) Desertification in Australia. In: Arid Land Development and the Combat againstDesertification: an Integrated Approach, pp. 101–12.United Nations Environment Programme, Moscow.

MacFarlane, G.R., Pulkownik, A. & Burchett, M.D.(2003) Accumulation and distribution of heavy metalsin the grey mangrove, Avicennia marina (Forsk.)Vierh.: biological indication potential. EnvironmentalPollution 123, 139–51.

MacGill, J.T. (1958) Map of coastal landforms of theworld. Geographical Revue 48, 402–5.

Machado, W., Silva-Filho, E.V., Oliveira, R.R., et al.(2002) Trace metal retention in mangrove ecosystemsin Guanabara Bay, SE Brazil. Marine Pollution Bulletin44, 1277–80.

Macintyre, I.G. (1985) Submarine cements – the peloidal question. In: Carbonate Cements (Eds N.Schneidermann & P.M. Harris), pp. 109–16. SpecialPublication 36, Society for Economic Paleontologistsand Mineralogists, Tulsa, OK.

Mackay, N. (1990) Understanding the Murray. In: TheMurray (Eds N. Mackay & D. Eastburn), pp. viii–xix.Murray-Darling Basin Commission, Canberra.

MacKenzie, A.B., Cook, G.T. & McDonald, P. (1999)Radionuclide distributions and particle size associations

ESD01 9/5/06 3:09 PM Page 409

410 REFERENCES

in Irish Sea surface sediments: implications for actinidedispersion. Journal of Environmental Radioactivity44, 275–96.

Mackey, D. & Paterson, S. (1982) Fugacity revisited.Environmental Science and Technology 16, 654–60.

Macklin, M.G. (1988) A Fluvial GeomorphologicalBased Evaluation of Contamination of the Tyne Basin,North-east England by Sediment-borne Heavy Metals.Unpublished Report to the Natural EnvironmentResearch Council, Swindon, 29 pp.

Macklin, M.G. (1992) Metal contaminated soils andsediment: a geographical perspective. In: Managing theHuman Impact on the Natural Environment: Patternsand Processes (Ed. M.D. Newson), pp. 172–95.Belhaven Press, London.

Macklin, M.G. (1996) Fluxes and storage of sediment-associated heavy metals in floodplain systems: assess-ment and river basin management issues at a time ofrapid environmental change. In: Floodplain Processes(Eds M.G. Anderson, D.E. Walling & P.D. Bates), pp. 441–60. Wiley, Chichester.

Macklin, M.G. & Lewin, J. (1989) Sediment transferand transformation of an alluvial valley floor: theRiver South Tyne, Northumbria, UK. Earth SurfaceProcesses and Landforms 14, 232–46.

Macklin, M.G. & Lewin, J. (1993) Holocene river alluviation in Britain. Zeitschrift für GeomorphologieSupplement-Band 88, 109–22.

Macklin, M.G. & Lewin, J. (2003) River sediments,great floods and centennial-scale Holocene climatechange. Journal of Quaternary Science 18, 101–5.

Macklin, M.G., Rumsby, B.T. & Newson, M.D. (1992)Historic overbank floods and vertical accretion offine-grained alluvium in the lower Tyne valley, northeast England. In: Dynamics of Gravel-bed Rivers,Proceedings of the Third International Workshop onGravel-bed Rivers (Eds P. Billi, R.D. Hey, P. Tacconi& C. Thorne), pp. 564–80. Wiley, Chichester.

Macklin, M.G., Ridgway, J., Passmore, D.G., et al.(1994) The use of overbank sediment for geochem-ical mapping and contamination assessment: results from selected English and Welsh floodplains. AppliedGeochemistry 9, 689–700.

Macklin, M.G., Brewer, P.A., Balteanu, D., et al. (2003)The longer term fate and environmental significanceof contaminant metals released by the January andMarch 2000 mining tailings dam failures in theMaramure County, upper Tisa Basin, Romania. AppliedGeochemistry 18, 241–57.

Macleod, C.L., Scrimshaw, M.D., Emmerson, R.H.C., et al. (1999) Geochemical changes in metal nutrientloading at Orplands Farm managed realignment site, Essex, UK. Marine Pollution Bulletin 38(12),1115–25.

Madrid, L., Diaz-Barrientos, E. & Madrid, F. (2002)Distribution of heavy metal contents of urban soils inparks of Seville. Chemosphere 49, 1301–8.

Magilligan, F.J., Gomez, B., Mertes, L.A.K., et al. (2002)Geomorphic effectiveness, sandur development, andthe pattern of landscape response during jökulhlaups:Skei¨oardrsandur, southeastern Iceland. Geomorpho-logy 44(1–2), 95–113.

Maizels, J. (1997) Jökulhlaup deposits in proglacialareas. Quaternary Science Reviews 16(7), 793–819.

Major, J.J. (2003) Post-eruption hydrology and sedimenttransport in volcanic river stystems. Water ResourcesImpact 5(3), 10–15.

Major, J.J., Pierson, T.C., Dinehart, R.L., et al. (2000)Sediment yield following severe volcanic disturbance – a two decade perspective from Mount St Helens.Geology 28, 819–22.

Major, J.J., Scott, W.E., Driedger, C., et al. (2005)Mount St. Helens Erupts Again: Activity fromSeptember 2004 through March 2005. Fact SheetFS2005-3036, U.S. Geological Survey, Denver, COhttp://vulcan.wr.usgs.gov/Volcanoes/MSH/Publications/FS2005-3036/FS2005-3036.pdf (accessed July 2005).

Mann, K.H. (2000) Ecology of Coastal Waters.Blackwell, Oxford.

Marsalek, J. & Marsalek, P.M. (1997) Characteristics of sediments from a stormwater management pond.Water Science Technology 36, 117–22.

Marston, R.A., Girel, J., Pautou, G., et al. (1995)Channel metamorphosis, floodplain disturbance, andvegetation development: Ain River, France. Geomor-phology 13, 121–31.

Martindale, W. (1992) Calcified epibionts as palaeoeco-logical tools: examples from the Recent and Pleistocenereefs of Barbados. Coral Reefs 11, 167–77.

Marui, H., Watanabe, N., Sato, O., et al. (1997) Gamaharatorrent debris flow on 6 December 1996, Japan – 10 debris flow disaster. Landslide News 10, 4–6.

Marutani, T., Brierley, G.J., Trustrum, N.A., et al. (2001)Source-to-sink Sedimentary Cascades in Pacific RimGeosystems. Matsumoto Sabo Works Office, Ministryof Land, Infrastructure and Transport, Japan.

Massadeh, A.M. & Snook, R.D. (2002) Determinationof Pb and Cd in road dusts over the period in which Pb was removed from petrol in the UK. Journal ofEnvironmental Monitoring 4, 567–72.

Masselink, G. & Short, A.D. (1993) The influence of tide range on beach morphodynamics: a conceptualmodel. Journal of Coastal Research 9, 785–800.

May, J.P. & Tanner, W.F. (1973) The littoral power gradient and shoreline changes. In: Coastal Geomor-phology (Ed. D.R. Coates), pp. 43–61. State Universityof New York, Binghampton.

May, R.M. (1978) Human reproduction reconsidered.Nature 272, 491–5.

McAlister, J.J., Smith, B.J. & Neto, J.A.B. (2000) Thepresence of calcium oxalate dihydrate (weddellite) in street sediments from Niteroi, Brazil and its healthimplications. Environmental Geochemistry and Health22, 195–210.

ESD01 9/5/06 3:09 PM Page 410

REFERENCES 411

McCarthy, J.J., Canziani, O.F., Leary, N.A., et al.(2001) Climate Change 2001: Impacts, Adaptationand Vulnerability. Contribution of Working Group IIto the Third Annual Assessment Report of the Inter-governmental Panel on Climate Change. CambridgeUniversity Press, Cambridge.

McCauley, J.E., Parr, R.A. & Hancock, D.R. (1977)Benthic infauna and maintenance dredging: a casestudy. Water Research 11, 233–42.

McCauley, J.F., Breed, C.S., El-Baz, F., et al. (1979)Pitted and fluted rocks in the western Desert of Egypt –Viking comparisons. Journal of Geophysical Research84, 8222–32.

McCauley, J.F., Ward, A.W., Breed, C.S., et al. (1977) Experimental Modelling of Wind ErosionForms. Technical Memorandum X-3511, NationalAeronautics and Space Administration, Washington,DC, pp. 150–2.

McCave, I.N. (1970) Deposition of fine-grained suspendedsediments from tidal currents. Journal of GeophysicsResearch 75, 4151–9.

McClanahan, T.R. (2000) Bleaching damage and recovery potential of Maldivian coral reefs. MarinePollution Bulletin 40, 587–97.

McCoy, E.D., Mushinsky, H.R., Johnson, D., et al.(1996) Mangrove damage caused by HurricaneAndrew on the southwestern coast of Florida. Bulletinof Marine Science 59, 1–8.

McFadden, L.D., Wells, S.G. & Jercinovich, M.J. (1987)Influences of aeolian and pedogenic processes on theorigin and evolution of desert pavements. Geology15, 504–8.

McFadden, L.D., Wells, S.G., Dohrenwend, J.C., et al.(1984) Cumulic soil formed in eolian parent materialson flows of the Cima Volcanic field, Mohave Desert.Geological Society of America Guidebook 14, AnnualMeeting, Reno, NV, pp. 134–49.

McGregor, G.R. & Nieuwolt, S. (1998) TropicalClimatology. Wiley, Chichester.

McKee, E.D. (Ed.) (1979) A Study of Global Sand Seas. Professional Paper 1052, US Geological Survey,Denver, CO.

McKenna-Neumann, C. & Nickling, W.G. (1989) Atheoretical and wind tunnel investigation of the effectof capillary water on the entrainment of soil by wind.Canadian Journal of Soil Science 69, 79–96.

McKnight, D.M. & Bencala, K.E. (1989) Reactive irontransport in an acidic mountain stream in SummitCounty, Colorado: a hydrologic perspective. Geo-chimica et Cosmochimica Acta 53, 2225–34.

McLusky, D.S. (1989) The Estuarine Ecosystem, 2ndedn. Blackie, Glasgow.

McMahon, T.A. (1979) Hydrological characteristics ofarid zones. In: The Hydrology of Areas of Low Pre-cipitation. Proceedings of a Symposium held duringthe XVII General Assembly of the IUGG at Canberra,December, pp. 105–23. IAHS Publication 128, Inter-

national Association of Hydrological Sciences Press,Wallingford.

McManus, J. (1988) Grain size determination and interpretation. In: Techniques in Sedimentology (Ed.M.E. Tucker), pp. 63–85. Blackwell Science, Oxford.

McSaveney, M.J. (1978). Sherman Glacier rock avalanche,Alaska, USA. In: Rockslides and Avalanches, Vol. 1,Natural Disasters (Ed. B. Voight), pp. 197–58. Elsevier,Amsterdam.

Meade, R.H. & Parker, R.S. (1985) Sediment in rivers of the United States. U.S. Geological Survey Water-supply Paper 2275, 49–60.

Mei-e, R. & Xianmo, Z. (1994) Anthropogenic influ-ences on changes in the sediment load of the YellowRiver, China, during the Holocene. The Holocene 4,314–20.

Meigs, P. (1953) World distribution of arid and semi-arid homoclimates. In: Arid Zone Hydrology.UNESCO Arid Zone Research Series 1, 203–9.

Memon, F.A. & Butler, D. (2002a) Identification andmodelling of dry weather processes in gully pots.Water Research 36, 1351–9.

Memon, F.A. & Butler, D. (2002b) Assessment of gullypot management strategies for runoff quality controlusing a dynamic model. The Science of the TotalEnvironment 295, 115–29.

Merilehto, K., Kenttämies, K. & Kämäri, J. (1988) Sur-face Water Acidification in the ECE Region. NordicCouncil of Ministers, NORD 1988:89, 156 pp.

Merrington, G. & Alloway, B.J. (1994) The transfer and fate of Cd, Cu, Pb and Zn from two historic metal-liferous mine sites in the U.K. Applied Geochemistry9, 677–87.

Messerli, B. (1983) Stability and instability of mountainecosystems: introduction to the workshop. MountainResearch and Development 3(2), 81–94.

Meybeck, M. (1979) Concentrations des eaux fluvialesen éléments majeurs et apports en solution aux océons.Revue de Géologie Dynamique et de GéographiePhysique 21, 215–46.

Meybeck, M. (2003) Global analysis of river systems:from Earth system controls to Anthropocene syn-dromes. Philosophical Transactions of the RoyalSociety of London, Series B, Online DOI 10.1098/rstb.2003.1979, 21 pp.

Middleton, N. (1997) Desert Dust. In: Arid ZoneGeomorphology: Process, Form and Change inDrylands, 2nd edn (Ed. D.S.G. Thomas), pp. 413–36.Wiley, Chichester.

Miguel, A.G., Cass, G.R., Glovsky, M.M., et al. (1999)Allergens in paved road dust and airborne particles.Environmental Science and Technology 33, 4159–68.

Miller, A.J. & Gupta, A. (Eds) (1999) Varieties ofFluvial Form. Wiley, Chichester.

Miller, C. (1999) Field measurements of longshore sedi-ment transport during storms. Coastal Engineering36, 301–21.

ESD01 9/5/06 3:09 PM Page 411

412 REFERENCES

Miller, J.R. (1997) The role of fluvial geomorphic pro-cesses in the dispersal of heavy metals from mine sites.Journal of Geochemical Exploration 58, 101–18.

Miller, J.R., Lechler, P.J. & Desilets, M. (1998) The roleof geomorphic processes in the transport and fate ofmercury within the Carson River Basin, west-centralNevada. Environmental Geology 33, 249–62.

Milliman, J.D. (2001) Delivery and fate of fluvial waterand sediment to the sea: a marine geologists’s view of European rivers. In: A Marine Science Odyssey into the 21st Century (Eds J.M. Gili, J.L. Pretus &T.T. Packard). Scientia Marina 65, 121–32.

Milliman, J.D. & Meade, R.H. (1983) World-widedelivery of river sediment to the oceans. Journal ofGeology 91, 1–21.

Milliman, J.D. & Syvitski, P.M. (1992) Geomorphic/tectonic control of sediment discharge to the ocean: theimportance of small mountainous terrain. The Journalof Geology 100, 525–44.

Milliman, J.D., Broadus, J.M. & Gable, F. (1989)Environmental and economic implications of risingsea level and subsiding deltas: the Nile and Bengalexamples. Ambio 18, 340–5.

Millington, A.C. (1989) African soil erosion: natureundone and the limits of technology. Land degrada-tion and Rehabilitation 1, 279–90.

Möeller, I., Spencer, T. & French, J.R. (1996) Windwave attenuation over salt marsh surfaces: preliminaryresults from Norfolk, England. Journal of CoastalResearch 12, 1009–16.

Mogi, A. (1979) An Atlas of the Sea Floor around Japan:Aspects of Submarine Geomorphology. University ofTokyo Press.

Monitor (1986) Acid and acidified waters (in Swedish).Swedish Environmental Protection Agency, Solna,180 pp.

Moore, J.N., Brook, E.J. & Johns, C. (1987) Grain sizepartitioning of metals in contaminated, coarse-grained floodplain sediment: Clark Fork River,Montana. Environmental Geology and Water Science14, 107–15.

Morehead, M.K., Syvitski, J.P. & Hutton, E.W.H.(2001) The link between abrupt climate change andbasin stratigraphy: a numerical approach. Global andPlanetary Change 28, 107–27.

Morgan, K.P.C. (1994) Soil Erosion and Conservation.Longman, London.

Morris, S.E. & Moses, T.A. (1987) Forest fire and thenatural soil erosion regime in the Colorado FrontRange. Association of American Geographers 77,245–54.

Morrison, G.M., Revitt, D.M., Ellis, J.B., et al. (1988)Transport mechanisms and processes for metal speciesin a gully pot system. Water Research 22, 1417–27.

Morrison, G.M., Revitt, D.M. & Ellis, J.B. (1995) Thegully pot as a biochemical reactor. Water ScienceTechnology 31, 229–36.

Morton, R.A., Gibeaut, J.C. & Paine, J.G. (1995) Meso-scale transfer of sand during and after storms:implications for prediction of shoreline movement.Marine Geology 126, 161–79.

Mosley, M.P. & Schumm, S.A. (2001) Gravel bed rivers– the view from the hills. In: Gravel-bed Rivers V(Ed. M.P. Mosley), pp. 479–505. The Caxton Press,Christchurch.

Motelica-Heino, M., Rauch, S., Morrison, G.M., et al.(2001) Determination of palladium, platinum andrhodium concentrations in urban road sediments by laser ablation-ICP-MS. Analytica Chimica Acta436, 233–44.

Mount, N.J., Sambrook Smith, G.H. & Stott, T.A. (2005)An assessment of the impact of upland afforestationon lowland river reaches: the Afon Trannon, mid-Wales. Geomorphology 64, 255–69.

Mualem, Y., Assouline, S. & Rohdenburg, H. (1990)Rainfall induced soil seal: a critical review of observa-tion and models. Catena 17, 185–203.

Mulder, J.P.M., van Koningsveld, M., Owen, M.W., et al. (2001) Guidelines on the Selection of CZMtools. Report RIKZ/2001.020, Rijkswaterstaat/RIKZ,UK Environment Agency, Swindon.

Muller, A. & Wessels, M. (1999) The flood in the Odra river 1997 – impact of suspended solids onwater quality. Acta Hydrochim. Hydrobiologia 27,316–20.

Munawar, M. & Dave, G. (Eds) (1996) Development andProgress in Sediment Quality Assessment: Rationale,Challenges, Techniques and Strategies. SPB AcademicPublishing, Amsterdam, 255 pp.

Munoz, D., Guiliano, M., Doumenq, P., et al. (1997)Long term evolution of petroleum biomarkers in man-grove soils (Guadeloupe). Marine Pollution Bulletin34, 868–74.

Myers, N. (1993) Gaia: an Atlas of Planet Management.Anchor and Doubleday, New York.

Nageotte, S.M. & Day, J.P. (1998) Lead concentra-tions and isotope ratios in street dust determined byelectrothermal atomic absorption spectrometry andinductively coupled plasma mass spectrometry. Analyst123, 59–62.

Nanayama, F., Shigeno, K., Satake, K., et al. (2000)Sedimentary differences between the 1993 Hokkaido-nanse-oki tsunami and the 1959 Miyakojima typhoonat Taisea, southwestern Hokkaido, northern Japan.Sedimentary Geology 135, 255–64.

Nash, D.J., Thomas, D.S.G. & Shaw, P.A.A. (1994)Timescales, environmental change and dryland valleydevelopment. In: Environmental Change in Drylands:Biogeographical and Geomorphological Perspectives(Eds A.C. Millington & K. Pye), pp. 25–41. Wiley,Chichester.

Neall, V.E. (1976) Lahars as major geological hazards.Bulletin of the International Association of Engineer-ing Geologists 14, 233–40.

ESD01 9/5/06 3:09 PM Page 412

REFERENCES 413

Needleman, H.L., Gunnoe, C. & Leviton, A. (1979)Deficits in psychological and classroom performancein children with elevated dentine lead levels. NewEngland Journal of Medicine 300, 689–95.

Neil, D.T., Orpin, A.R., Ridd, P.V., et al. (2002)Sediment yield and impacts from river catchments tothe Great Barrier Reef lagoon. Marine and FreshwaterResearch 53, 733–52.

Nelsen, J.E. & Ginsburg, R.N. (1986) Calcium carbonateproduction by epibionts on Thalassia in Florida Bay.Journal of Sedimentary Petrology 56, 622–8.

Nelson, E.J. & Booth, D.B. (2002) Sediment sources inan urbanizing, mixed land-use watershed. Journal ofHydrology 264, 51–68.

Nelson-Smith, A. (1972) Effects of oil industry on shorelife in estuaries. Proceedings of the Royal Society ofLondon 180B, 487–96.

Nemec, W. & Steel, R.J. (1984) Alluvial and coastal conglomerates: their significant features and somecomments on gravely mass-flow deposits. In: Sedimento-logy of Gravels and Conglomerates (Eds E.H. Koster& R.J. Steel), pp. 1–31. Memoir 10, Canadian Societyof Petroleum Geologists, Calgary.

Neumann, A.C. (1966) Observations on coastal erosionin Bermuda and measurement of the boring rate of thesponge, Cliona lampa. Limnology and Oceanography11, 92–108.

Neumann, A.C. & Land, L.S. (1975) Lime mud deposi-tion and calcareous algae in the Bight of Abaco,Bahamas: a budget. Journal of Sedimentary Petrology45, 763–86.

Neumann, A.C. & Macintyre, I. (1985) Reef response tosea-level rise: keep-up, catch-up or give-up. Proceed-ings of 5th International Coral Reef Symposium, Vol. 3,pp. 105–10. National Museum of Natural Historyand the Practical School of Advanced Studies, Tahiti,French Polynesia.

Newson, M. (1992) Land, Water and Development. RiverBasin Systems and their Sustainable Management.Routledge, London.

Newson, M.D. (1980) The erosion of drainage ditchesand its effects on bed load yields in Wales: recon-naissance case studies. Earth Surface Processes andLandforms 5, 275–90.

Newton, R.S., Siebold, E. & Werner, F. (1973) Faciesdistribution patterns on the Spanish Saharan con-tinental shelf mapped with side-scan sonar. ‘Meteor’Forschungsergebnisse, Reihe C: Geologie undGeophysik 15, 55–77.

Nguyen, V.L., Oanh, T.K. & Tateishi, M. (2000) LateHolocene depositional environments and coastal evo-lution of the Mekong River Delta, southern Vietnam.Journal of Asian Earth Sciences 18, 427–39.

Nicholls, R.J. (2004) Coastal flooding and wetland lossin the 21st Century: changes under the SRES climateand socio-economic scenarios. Global EnvironmentalChange 14, 69–86.

Nickling, W.G. (1988) The initiation of particle move-ment by wind. Sedimentology 35, 499–511.

Nickling, W.G. (1994) Aeolian sediment transport and deposition. In: Sediment Transport and Deposi-tional Processes (Ed. K. Pye), pp. 293–350. Blackwell,Oxford.

Nittrouer, C.A. & Wright, L.D. (1994) Transport ofparticles across continental shelves. Reviews in Geo-physics 32, 85–113.

Nittrouer, C.A., Sternberg, R.W., Carpenter, R., et al.(1979) The use of 210Pb geochronology as a sedi-mentological tool: application to the Washington continental shelf. Marine Geology 31, 297–316.

Noormets, R., Felton, E.A. & Crook, K.A.W. (2002)Sedimentology of rocky shorelines: 2 Shoreline mega-clasts on the north shore of Oahu, Hawaii – originsand history. Sedimentary Geology 150, 31–45.

Nordstrom, K.F. (2000) Beaches and Dunes ofDeveloped Coasts. Cambridge University Press, Cam-bridge, 383 pp.

Novotny, V. (2003) Water Quality, Diffuse Pollution andWatershed Management, 2nd edn. Wiley, New York.

Nürnberg, G.K. (1996) Trophic state of clear and colored,soft- and hardwater lakes with special considerationof nutrients, anoxia, phytoplankton and fish. Journalof Lake and Reservoir Management 12, 432–47.

Nürnberg, G.K. & Shaw, M. (1998) Productivity ofclear and humic lakes: nutrients, phytoplankton, bac-teria. Hydrobiologia 382, 97–112.

O’Hara, S.L., Street-Perrott, F.A. & Burt, T.P. (1993)Accelerated soils erosion around a Mexican highlandlake caused by pre-hispanic agriculture. Nature 363,48–51.

Oak, H.L. (1984) The boulder beach: a fundamentallydistinct sedimentary assemblage. Annals of the Asso-ciation of American Geographers 74, 71–82.

Oberlander, T.M. (1994). Rock varnish in deserts. In:Geomorphology of Desert Environments (Eds A.D.Abrahams & A.J. Parsons), pp. 107–19. Chapman &Hall, London.

Odin, G.S. (1988) Green marine clays: oolitic ironstonefacies, verdine facies, glaucony facies, and celadonite-bearing facies: a comparative study. Developments inSedimentology 45. Elsevier, New York, 445 pp.

OECD (1982) Eutrophication of Waters. Monitoring,Assessment and Control. Organization of EconomicCooperation and Development, Paris, 154 pp.

OECD (1991) Environmental Indicators. Organizationof Economic Cooperation and Development, Paris, 77 pp.

Oertel, G.F. (1985) The barrier island system. MarineGeology 63, 1–18.

Old, G.H., Leeks, G.J.L., Packman, J.C., et al. (2003)The impact of a convectional summer rainfall event on river flow and fine sediment transport in a highlyurbanised catchment: Bradford, West Yorkshire. TheScience of the Total Environment 314–16, 495–512.

ESD01 9/5/06 3:09 PM Page 413

414 REFERENCES

Oldfield, F., Hunt, A., Jones, M.D.H., et al. (1985)Magnetic differentiation of atmospheric dusts. Water,Air and Soil Pollution 317, 516–18.

O’Neill, R.V., Gardner, R.H., Barnthouse, L.W., et al.(1982) Ecosystem risk analysis: a new methodo-logy. Environmental Toxicology and Chemistry 1,167–77.

Orford, J.D., Carter, R.W.G., McKenna, J., et al. (1995)The relationship between the rate of mesoscale sea-level rise and the retreat rate of swash-alignedgravel-dominated coastal barriers. Marine Geology124, 177–86.

Orford, J.D., Cooper, J.A.G. & McKenna, J. (1999)Mesoscale temporal changes to foredunes on Inch Spit,south-west Ireland. Zeitschrift fur GeomorphologieN.F. 43, 439–61.

Orford, J.D., Murdy, J.M. & Wintle, A.G. (2003)Prograded Holocene beach ridges with superimposeddunes in north-east Ireland: mechanisms and time-scales of fine and coarse beach sediment decouplingand deposition. Marine Geology 194, 47–64.

Ornstein, P., Jaboyedoff, M. & Rouiller, J.D. (2001)Surveillance géodésique du site de Randa (VS): gestiondes measures 1-D et 3-D. Publication Soc, Suisse MécSols Roches 143, 82–91.

Orpin, A.R., Ridd, P.V. & Stewart, L.K. (1999) Assess-ment of the relative importance of major sedimenttransport mechanisms in the central Great BarrierReef lagoon. Australian Journal of Earth Sciences 46,883–96.

OSPAR Commission (2000) Quality Status Report2000, Region II – Greater North Sea. OSPAR (Oslo/Paris convention (for the Protection of the MarineEnvironment of the North-East Atlantic)) Commis-sion, London, 136 + xiii pp. (www.ospar.org)

Ostapenia, A.P. (1985) Ratio between the componentsof seston. Ecological system of Naroch lakes. Minsk,pp. 232–3 (in Russian).

Ouchi, S. (1985) Response of alluvial rivers to slow activetectonic movement. Geological Society of AmericaBulletin 96, 504–15.

Owens, P.N. & Slaymaker, O. (Eds) (2004) MountainGeomorphology. Arnold, London, 313 pp.

Owens, P.N. & Walling, D.E. (2002) The phosphoruscontent of fluvial sediment in rural and industrializedriver basins. Water Research 36, 685–701.

Owens, P.N., Walling, D.E. & Leeks, G.J.L. (2000)Tracing fluvial suspended sediment sources in thecatchment of the River Tweed, Scotland, using com-posite fingerprints and a numerical mixing model. In: Tracers in Geomorphology (Ed. I.D.L. Foster), pp. 291–308. Wiley, Chichester.

Owens, P.N., Walling, D.E., Carton, J., et al. (2001)Downstream changes in the transport and storage ofsediment-associated contaminants (P, Cr and PCBs) in agricultural and industrialized drainage basins. The Science of the Total Environment 266, 177–86.

Owens, P.N., Apitz, S., Batalla, R., et al. (2004) Sedi-ment management at the river basin scale: synthesis ofthe SedNet Working Package 2 outcomes. Journal ofSoils and Sediments 4, 219–22.

Page, M.J., Trustrum, N.A. & Dymond, J.R. (1994)Sediment budget to assess the geomorphic effect of acyclonic storm, New Zealand. Geomorphology 9(3),169–88.

Page, M.J., Reid, L.M. & Lynn, I.H. (1999) Sedimentproduction from Cyclone Bola landslides, Waipaoacatchment. Journal of Hydrology (New Zealand)38(2), 289–308.

Painter, R.B., Blyth, K., Mosedale, J.C. & Kelly, M.(1974) The effects of afforestation on erosion pro-cesses and sediment yields. In: Effects of Man on theInterface of the Hydrological Cycle with the PhysicalEnvironment, Proceedings of a Symposium held atParis, September, pp. 62–7. IAHS Publication 113,International Association of Hydrological SciencesPress, Wallingford.

Palanques, A., Guilen, J. & Puig, P. (2001) Impact ofbottom trawling on water turbidity and muddy sedi-ment of an unfished continental shelf. Limnology andOceanography 46, 1100–10.

Palermo, M.R., Clausner, J.E., Rollings, M.P., et al.(1998) Guidance for Sub-aqueous Dredged MaterialCapping. Technical Report DOER-1, U.S. ArmyEngineer Waterways Experiment Station, Vicksburg,MS.

Parkman, R.H., Curtis, C.D., Vaughan, D.J. &Charnock, J.M. (1996) Metal fixation and mobiliza-tion in the sediments of Afon Goch Estuary, DulasBay, Anglesey. Applied Geochemistry 11, 203–10.

Passmore, D.G. & Macklin, M.G. (1994) Provenance of fine-grained alluvium and late Holocene land-usechange in the Tyne basin, northern England. Geomor-phology 9, 127–42.

Pearce, F. (1996) Crumbling away. New Scientist 152,1061–2.

Pearson, T.H. & Rosenberg, R. (1976) A comparativestudy on the effects on the marine environment ofwastes from cellulose industries in Scotland andSweden. Ambio 5, 77–9.

Peart, M.R. & Walling, D.E. (1986) Fingerprinting sediment source: the example of a drainage basin inDevon, UK. In: Drainage Basin Sediment Delivery,pp. 41–55. IAHS Publication 159, International Asso-ciation of Hydrological Sciences Press, Wallingford.

Pedley, H.M. (1990) Classification and environmentalmodels of cool freshwater tufas. Sedimentary Geology68, 143–54.

Pedley, H.M. (1992) Freshwater (phytoherm) reefs: the role of biofilms and their bearing on marine reefcementation. Sedimentary Geology 79, 255–74.

Pedley, M. (2000) Ambient temperature freshwater micro-bial tufas. In: Microbial Sediments (Eds R.E. Riding &S.M. Awramik). Springer-Verlag, Berlin.

ESD01 9/5/06 3:09 PM Page 414

REFERENCES 415

Pedley, M., Andrews, J. & Ordonez, S. (1996) Does climate control the morphological fabric of freshwatercarbonates? A comparative study of Holocene barragetufas from Spain and Britian. Palaoegeography, Palao-eclimatology, Palaoeecology 121, 239–57.

Peel, R.A. (1970) Landscape sculpture by wind. 21stInternational Geographical Congress Selected Papers1, 99–104.

Pelc, R. & Fujita, R.M. (2002) Renewable energy fromthe ocean. Marine Policy 26, 471–9.

Pennington, W., Cambray, R.S. & Fischer, E.M. (1973)Observations on lake sediments using fallout of Cs-137 as a tracer. Nature 242, 324–6.

Perry, C.T. (1996) The rapid response of reef sedimentsto changes in community structure: implications fortime-averaging and sediment accumulation. Journalof Sedimentary Research 66, 459–67.

Perry, C.T. (1998a) Macroborers within coral frameworkat Discovery Bay, north Jamaica: species distributionand abundance and effects on coral preservation.Coral Reefs 17, 277–87.

Perry, C.T. (1998b) Grain susceptibility to the effects of microboring: implications for the preservation ofskeletal carbonates. Sedimentology 45, 39–51.

Perry, C.T. (1999). Reef framework preservation in fourcontrasting modern reef environments, Discovery Bay,Jamaica. Journal of Coastal Research 15, 796–812.

Perry, C.T. (2000) Factors controlling sediment preser-vation on a north Jamaican fringing reef: a processbased approach to microfacies analysis. Journal ofSedimentary Research 70, 633–48.

Perry, C.T. (2003) Coral reefs in a high latitude, siliciclastic barrier island setting: reef framework and sediment production at Inhaca Island, southernMozambique. Coral Reefs 22, 485–97.

Perry, C.T. (2005) Structure and development of detritalreef deposits in turbid nearshore environments, InhacaIsland, Mozambique. Marine Geology 214, 143–61.

Perry, C.T. & Beavington-Penney, S.J. (2005) Epiphyticcalcium carbonate production and facies develop-ment from seagrass beds in high latitude, reef-relatedenvironments, Inhaca Island, Mozambique. Sedi-mentary Geology 174, 161–76.

Perry, C.T. & Larcombe, P. (2003) Marginal and non-reef building coral environments. Coral Reefs 22, 427–32.

Perry, C.T. & Taylor, K.G. (2004) Impacts of bauxitesediment inputs on a carbonate-dominated embay-ment, Discovery Bay, Jamaica. Journal of CoastalResearch 20, 1070–9.

Pethick, J. (1984) An Introduction to Coastal Geomor-phology. Edward Arnold, London.

Petit-Maire, N. & Guo, Z.T. (1998) Mid-Holocene climatic change and Man in the present-day Saharadesert. In: Quaternary Deserts and Climatic Change(Eds A.S. Alsharhan, K.W. Glennie, G.L. Whittle &C.G.St.C. Kendall), pp. 351–6. Balkema, Rotterdam.

Pettijohn, F.J., Potter, P.E. & Siever, R. (1987) Sand andSandstone. Springer-Verlag, Berlin, 553 pp.

Petts, G.E. (1984) Impounded Rivers. Wiley, Chichester.Piégay, H., Walling, D.E., Landon, N., et al. (2004)

Contemporary changes in sediment yield in an alpinemountain basin due to aforestation (the upper Drômein France). Catena 55, 183–212.

Pierson, T.C. (1988) Hazardous hydrologic consequencesof volcanic eruptions and goals for mitigative action:an overview. In: Hydrology of Disasters Proceed-ings Technical Conference WNMO, Geneva (Eds O. Starosolszky & O.M. Melder), pp. 220–36.

Pilkey, O.H. (2003) Barrier Islands. Columbia Univer-sity Press, Columbia, OH.

Pilkey, O.H. & Cooper, J.A.G. (2004) Society and sealevel rise. Science 303, 1781–2.

Pilkey, O.H. & Dixon, K.L. (1996) The Corps and theShore. Island Press, Washington, DC.

Pilkey, O.H., Neal, W.J., Riggs, S.R., et al. (1998) TheNorth Carolina Shore and its Barrier Islands. DukeUniversity Press, Durham, NC.

Pirazzoli, P.A. (1996) Sea-level Change: the Last 20,000Years. Wiley, Chichester.

Pirrie, D., Power, M.R., Wheeler, P.D. & Ball, A.S.(2000) A new occurrence of diagenetic simonkolleitefrom the Ganel Estuary, Cornwall. Geoscience inSouthwest England 10, 18–20.

Plafker, G. & Eriksen, G.E. (1978) Nevados Huascaranavalanches, Peru. In: Rockslides and Avalanches, Vol. 1, Natural Disasters (Ed. B. Voight), pp. 48–55.Elsevier, Amsterdam.

Plaziat, J.C. (1974) Mollusc distribution and its valuefor recognition of ancient mangroves. Proceedings of International Symposium on the Biology andManagement of Mangroves, Honolulu, Vol. 2, pp. 456–65.

Plaziat, J.C. (1995) Modern and fossil mangroves andmangals: their climatic and biogeographic variability.In: Marine Palaeoenvironmental Analysis from Fossils(Eds D.W.J. Bosence & P.A. Allison), pp. 73–96.Special Publication 83, Geological Society PublishingHouse, Bath.

Poesen, J. (1986) Surface sealing as influenced by slopeangle and position of simulated stones in the top layer of loose sediments. Earth Surface Processes andLandforms 11, 1–10.

Porter-Smith, R., Harris, P.T., Andersen, O.B., et al.(2004) Classification of the Australian continentalshelf based on predicted sediment threshold exceed-ance from tidal currents and swell waves. MarineGeology 211, 1– 20.

Postma, G. (1986) Classification for sediment gravityflow deposits based on flow conditions during sedi-mentation. Geology 14, 291–4.

Potts, D.C. & Jacobs, J.R. (2003) Evolution of reef-building scleractinian corals in turbid environments: a paleo-ecological hypothesis. In: Proceedings of the

ESD01 9/5/06 3:09 PM Page 415

416 REFERENCES

9th International Coral Reef Symposium, Bali, Vol. 2,pp. 249–54. Ministry of Environment (Indonesia),Indonesian Institute of Sciences and InternationalSociety for Reef Studies.

Power, J, McKenna, J., MacLeod, M., et al. (2000)Developing integrated participatory managementstrategies for Atlantic dune systems in County Donegal,Ireland. Ambio 29, 143–9.

Powers, M.C. (1982) Comparison Chart for Estimat-ing Roundness and Sphericity. AGI Data Sheet 18,American Geological Institute, Alexandria, VA.

Preu, C. (1989) Coastal erosion in southwestern Sri Lanka:consequences of human interference. Malaysian Journalof Tropical Geography 20, 30–42.

Price, L.W. (1981) Mountain and Man: a study ofProcess and Environment. University of CaliforniaPress, Berkeley.

Pringle, A.W. (1995) Erosion of a cyclic saltmarsh inMorecambe Bay, North-West England. Earth SurfaceProcesses and Landforms 20, 387–405.

Productivity Commission (2003) Industries, Land Useand Water Quality in the Great Barrier Reef Catch-ment. Research Report, 415 pp. (available at http://www.pc.gov.au/study/gbr/finalreport/index.html).

Prospero, J.M., Bonatti, E., Schubert, C., et al. (1970) Dustin the Caribbean atmosphere traced to an African duststorm. Earth and Planetary Science Letters 9, 287–93.

Purdy, E.G. (1963) Recent calcium carbonate facies ofthe Great Bahama Bank. 2. Sedimentary facies.Journal of Sedimentary Petrology 71, 474–97.

Pye, K. & French, P.W. (1993) Targets for CoastalHabitat Re-creation. English Nature Science SeriesNo. 13, English Nature, Peterborough.

Pye, K. & Tsoar, H. (1990) Aeolian Sand and SandDunes. Unwin Hyman, London. 396 pp.

Qu, W. & Kelderman, P. (2001) Heavy metal contents inthe Delft canal sediments and suspended solids of theRiver Rhine: multivariate analysis for source tracing.Chemosphere 45, 919–25.

Quanterra (2003) Short Guide about Slope Instabil-itites between Lausanne and Zermatt (Switzerland).(18) Randa. http://www.quanterra.org/guide/guide1_18.htm. Accessed May, 2005.

Radoane, M., Radoane, N. & Dumitriu, D. (2003) Geo-morphological evolution of longitudinal river profilesin the Carpathians. Geomorphology 50, 293–306.

Rae, J.E. & Allen, J.R.L. (1993) The significance oforganic matter degradation in the interpretation ofhistorical pollution trends in depth profiles of estuarinesediment. Estuaries 16, 678–82.

Raetzo, H., Latelin, O., Bollinger, D., et al. (2002)Hazard assessment in Switzerland – codes of practicefor mass movements. Bulletin of Engineering Geologyand the Environment 61, 263–8.

Rahn, K.A., Borys, R.D. & Shaw, G.E. (1981) Asiandesert dust over Alaska: anatomy of an Arctic haze

episode. In: Desert Dust (Ed. T.L. Péwé), pp. 37–70.Special Paper 186, Geological Society of America,Boulder, CO.

Ramsay, M.A., Swannells, R.P.J., Shipton, W.A., et al.(2000) Effect of bioremediation on the microbial community in oiled mangrove sediments. MarinePollution Bulletin 41, 413–19.

Rapp, A. (1960) Recent developments of mountain slopesin Kärkevagge and surroundings, northern Scandanavia.Geografiska Annaler 42A, 71–200.

Rasser, M.W. & Riegl, B. (2002) Holocene coral reefrubble and its binding agents. Coral Reefs 21, 57–72.

Rawat, J.S. & Rawat, M.S. (1994) Accelerated erosionand denudation in the Nani Kosi watershed, CentralHimalaya, India. Part I: sediment load. MountainResearch and Development 14, 25–38.

Reading, H.G. (Ed.) (1996) Sedimentary Environments:Processes, Facies and Stratigraphy, 3rd edn. BlackwellScientific, Oxford, 688 pp.

Reddering, J.S.V. (1983) An inlet sequence produced bymigration of a small microtidal inlet against longshoredrift: the Keurbooms inlet, South Africa. Sedimentology30, 201–18.

Reeves, C.C., Jr. (1968) Introduction to Paleolimnology.Elsevier, Amsterdam, 228 pp.

Reichett-Brushett, A.J. & Harrison, P.L. (1999) Theeffects of copper on the settlement success of larvaefrom the scleractinian coral Acropora tenuis. MarinePollution Bulletin 41, 385–91.

Reid, I. & Frostick, L.E. (1987) Flow dynamics and suspended sediment properties in arid zone flashfloods. Hydrological Processes 1, 239–53.

Reid, I. & Frostick, L.E. (1994) Fluvial sediment trans-port and deposition In: Sediment Transport andDepositional Processes (Ed. K. Pye), pp. 89–155.Blackwell Science, Oxford.

Reid, I. & Laronne, J.B. (1995) Bedload sediment trans-port in an ephemeral stream and a comparison withseasonal and perennial counterparts. Water ResourcesResearch 31, 773–81.

Reid, I., Laronne, J.B., Powell, D.M., et al. (1994). Flashfloods in desert rivers: studying the unexpected. Eos(Transactions of the American Geophysical Union)75, 452.

Reisner, M. (1986) Cadillac Desert: the American West and its disappearing water. Viking Penguin,London.

Renard, K.G. & Keppel, R.V. (1966) Hydrographs ofephemeral streams in the Southwest. Proceedings ofthe American Society of Civil Engineers, Journal of theHydraulics Division 92 (HY2), 33–52.

Rhoades, J.D. (1990) Soil salinity – causes and con-trols. In: Techniques for Desert Reclamation (Ed. A.S. Goudie), pp. 109–34. Wiley, Chichester.

Rice, S.P. & Church, M. (1998) Grain size along twogravel-bed rivers: statistical variation, spatial pattern

ESD01 9/5/06 3:09 PM Page 416

REFERENCES 417

and sedimetary links. Earth Surface Processes andLandforms 23, 345–63.

Richard, G.A. (1978) Seasonal and environmental variations in sediment accretion in a Long Island saltmarsh. Estuaries 1, 29–35.

Richards, K.S. (1982) Rivers: Form and Process inAlluvial Channels. London, Methuen.

Rickenmann, D. (1990) Debris flows 1987 in Switzerland:modelling and fluvial sediment transport. In: Hydro-logy in Mountainous Regions. II – Artificial Reservoirs;Water and Slopes (Eds R.O. Sinniger & M. Monbaron),pp. 371–8. IAHS Publication 194, Proceedings of two Symposia held at Lausanne, August. Interna-tional Association of Hydrological Sciences Press,Wallingford.

Rickenmann, D. & Zimmermann, M. (1993) The 1987debris flows in Switzerland: Documentation and analysis. Geomorphology 8, 175–89.

Ridd, P.V. (1996) Flow through animal burrows in mangrove creeks. Estuarine, Coastal and Shelf Science43, 617–25.

Ridd, P.V., Stieglitz, T. & Larcombe, P. (1998) Density-driven secondary circulation in a tropical mangroveestuary. Estuarine, Coastal and Shelf Science 47,621–32.

Ridge, M.J.H. & Carson, B. (1987) Sediment transporton the Washington continental shelf: estimates of dis-persal rates from Mount St. Helen’s ash. ContinentalShelf Research 7, 759–72.

Riegl, B. & Piller, W.E. (2000) Reefs and coral carpets in the northern Red Sea as models for organism–environment feedback in coral communities and itsreflection in growth fabrics. In: Carbonate PlatformSystems: Components and Interactions (Eds E. Insalaco,P.W. Skelton & T.J. Palmer), pp. 71–88. Special Pub-lication 178, Geological Society Publishing House,Bath.

Riegl, B., Schleyer, M.H., Cook, P.J., et al. (1995)Structure of Africa’s southernmost coral communities.Bulletin of Marine Science 56, 648–63.

Riethmüller, R., Hakvoort, J.H.M., Heineke, M., et al.(1998) Relating erosion shear stress to tidal flat sur-face colour. In: Sedimentary Processes in the IntertidalZone (Eds K.S. Black, D.H. Paterson & A. Cramp),pp. 283–93. Special Publication 139, GeologicalSociety Publishing House, Bath.

Rijnsdorp, A.D., Bujiis, A.M., Storbeck, F., et al. (1998)Microscale distribution of beam trawl effort in thesouthern North Sea between 1993 and 1996 in rela-tion to the trawling frequency of the sea bed and thedistribution of benthic organisms. ICES Journal ofMarine Science 55, 403–19.

Ringrose-Voase, A.J., Rhoades, D.W. & Hall, G.F.(1989) Reclamation of a scalded, red duplex soil bywaterponding. Australian Journal of Soil Research27, 779–95.

Rivera-Duarte, I. & Flegal, A.R. (1997a) Porewater gradients and diffusive benthic fluxes of Co, Ni, Cu,Zn and Cd in San Francisco Bay. Croatica ChemicaActa 70, 389–417.

Rivera-Duarte, I. & Flegal, A.R. (1997b) Pore-water silver concentrations and benthic fluxes from con-taminated sediments of San Francisco Bay, California,U.S.A. Marine Chemistry 56, 15–26.

Robbins, J.A. (1978) Geochemical and GeophysicalApplications of Radioactive Lead. In: Biogeochemistryof Lead in the Environment (Ed. J.O. Nriagu), pp. 285–93. Elsevier, Amsterdam.

Roberts, H.H., Wiseman, J. & Suchanek, T.H. (1981)Lagoon sediment transport: the significant effect ofCallianassa bioturbation. Proceedings of Fourth Inter-national Coral Reef Symposium, Vol. 1, pp. 459–65.Marine Science Center, University of the Philippines,Manila, Philippines.

Robertson, A.I., Alongi, D.M. & Boto, K.G. (1992)Food chains and carbon fluxes. In: Tropical MangroveEcosystems (Eds A.I. Robertson & D.M. Alongi), pp. 293–326. Coastal and Estuarine Studies 41,American Geophysical Union, Washington.

Robertson, D.J., Taylor, K.G. & Hoon, S.R. (2003)Geochemical and mineral characterisation of urbansediment particulates, Manchester. UK. Applied Geo-chemistry 18, 269–82.

Robinson, J.E., Newell, R.C., Seiderer, L.J., et al. (2005)Impacts of aggregate dredging on sediment composi-tion and associated benthic fauna at an offshore dredgesite in the southern North Sea. Marine EnvironmentalResearch 60, 51–68.

Robinson, M. & Lambrick, G.H. (1984) Holocene alluviation and hydrology in the Upper Thames basin.Nature 308, 809–14.

Rodier, M. (1992) The Rance tidal power station: aquarter of a century. In: Tidal Power: Trends andDevelopments (Eds Institute of Civil Engineers), pp. 301–8. Thomas Telford Press, London.

Roesli, U. & Schindler, C. (1990) Debris flows 1987 inSwitzerland: geological and hydrogeological aspects.In: Hydrology in Mountainous Regions. II – ArtificialReservoirs; Water and Slopes (Eds R.O. Sinniger &M. Monbaron), pp. 379–86. IAHS Publication 194,Proceedings of two Symposia held at Lausanne, August.International Association of Hydrological SciencesPress, Wallingford.

Rogers, C.S. (1990) Responses of coral reefs and reeforganisms to sedimentation. Marine Ecology ProgressSeries 62, 185–202.

Rogers, C.S. (1993) Hurricanes and coral reefs: the intermediate disturbance hypothesis revisited. CoralReefs 12, 127–37.

Rogers, J. & Li, X.C. (2002) Environmental impact ofdiamond mining on continental shelf sediments offsouthern Namibia. Quaternary International 92, 101–12.

ESD01 9/5/06 3:09 PM Page 417

418 REFERENCES

Rogers, S.I. & Greenaway, B. (2005) A UK perspectiveon the development of marine ecosystem indicators.Marine Pollution Bulletin 50, 9–19.

Rosen, B.R. (1990) Reefs and carbonate build-ups. In: Palaeobiology – a Synthesis (Eds D.E.G. Briggs & P.R. Crowther), pp. 341–6. Blackwell Scientific,Oxford.

Rosen, M.R. (1994) The importance of groundwater in playas: a review of playa classifications and the sedimentology and hydrology of playas. In: Paleo-climate and Basin Evolution of Playa Systems (Ed.M.R. Rosen), pp. 1–18. Special Paper 289, GeologicalSociety of America, Boulder, CO.

Rostad, C.E., Pereira, W.E. & Leiker, T.J. (1999)Distribution and transport of selected anthropo-genic lipophilic organic compounds associated withMississippi River suspended sediment, 1989–1990.Archives of Environmental Contamination andToxicology 36, 248–55.

Rowan, J.S., Goodwill, P. & Franks, S.W. (2000)Uncertainty estimation in fingerprinting suspendedsediment sources. In: Tracers in Geomorphology(Ed. I.D.L. Foster), pp. 279–90. Wiley, Chichester.

Rowlatt, S., Matthiessen, P., Reed, J., et al. (2002)Review and Recommendations of Methodologies forthe Derivation of Sediment Quality Guidelines. R&DTechnical Report P2-082/TR, Environment Agency,Bristol, 113 pp.

Roy, P.S., Cowell, P.J., Ferland, M.A., et al. (1994)Wave-dominated coasts. In: Coastal Evolution (EdsR.W.G. Carter & C.D. Woodroffe), pp. 121–86.Cambridge University Press, Cambridge.

Rozan, R.F., Lassman, M.E., Ridge, D.P., et al. (2000)Evidence for iron, copper and zinc complexation asmultinuclear sulphide clusters in oxic rivers. Nature406, 879–82.

Rumsby, B.T. & Macklin, M.G. (1994) Channel andfloodplain response to recent abrupt climate change:the Tyne basin, northern England. Earth SurfaceProcesses and Landforms 19, 499–515.

Rust, B.R. (1978) A classification of alluvial channel systems. In: Fluvial Sedimentology (Ed. A.D. Miall), pp. 187–98. Memoir 5, Canadian Society of PetroleumGeologists, Calgary.

Rust, B.R. & Waslenchuk, D.G. (1974) The distributionand transport of bed sediments and persistent pollu-tants in the Ottawa River, Canada. In: Proceedings ofthe International Conference on the Transportation of Persistent Chemicals in the Aquatic Ecosystem.Ottawa, Canada, pp. 1–25.

Ryding, S.O. & Borg, H. (1973) Sediment-chemicalStudies in Lake Lilla Ullevifjärden. Report 58, NationalSwedish Environmental Protection Board, NLU,Uppsala, 77 pp. (In Swedish.)

Saad, S., Husain, M.L., Yaacob, R., et al. (1999)Sediment accretion and variability of sedimentolo-gical characteristics of a tropical estuarine mangrove:

Kemaman, Terengganu, Malaysia. Mangroves andSalt Marshes 3, 51–8.

Saito, Y., Yang, Z. & Hori, K. (2001) The Huanghe(Yellow River) and Changjiang (Yangtze River)deltas: a review on their characteristics, evolution andsediment discharge during the Holocene. Geomor-phology 41, 219–31.

Salomons, W. & De Groot, A.J. (1978) Pollution historyof trace elements in sediments, as affected by theRhine River. In: Environmental Biochemistry, Vol. 1(Ed. W.E. Krumbein), pp. 149–62. Ann Arbor SciencePublications, Maine.

Salomons, W. & Förstner, U. (1984) Metals in theHydrocycle. Springer-Verlag, Heidelberg, 349 pp.

Sanderson, G. & Elliot, I.G. (1996) Shoreline salients,cuspate forelands and tombolos on the coast of WesternAustralia. Journal of Coastal Research 12, 761–73.

Santschi, P.H. & Honeyman, B.D. (1991) Radio-isotopes as tracers for the interactions between traceelements, colloids and particles in natural waters. In:Heavy Metals in the Environment (Ed. J.-P. Vernet),pp. 229–46. Elsevier, Amsterdam.

Sarnthein, M. & Walger, E. (1974) Der äolische sand-strom aus der W. Sahara zur Atlantikküste. GeologischeRundschau 63, 1065–87.

Sarnthein, M., Kennett, J.P., Allen, J.R.M., et al. &SCOR-IMAGES Working Group 117 (2002) Decadal-to-millennial-scale climate variability – chronologyand mechanisms: summary and recommendations.Quaternary Science Reviews 10, 1121–8.

Sarre, R.D. (1989) Aeolian sand drift from the intertidalzone on a temperate beach: potential and actual rates.Earth Surface Processes and Landforms 14, 247–58.

Sarre, R.D. (1990) Evaluation of aeolian sand trans-port equations using intertidal-zone measurements,Saunton Sands, England – reply. Sedimentology 37,389–92.

Sartor, J.D. & Gaboury, D.R. (1984) Street sweeping as a water pollution control measure: lessons learnedover the past ten years. The Science of the TotalEnvironment 33, 171–83.

Sartor, J.D., Boyd, G.B. & Agardy, F.J. (1974) Waterpollution aspects of street surface contaminants.Journal of Water Pollution Control Federation 46,458–67.

Schell, C., Black, S. & Hudson-Edwards, K.A. (2000)Sediment source characteristics of the Río Tinto, Huelva,south-west Spain. In: Tracers in Geomorphology (Ed.I.D.L. Foster), pp. 503–20. Wiley, Chichester.

Schick, A.P. (1970) Desert floods. In: Symposium of theResults of Research on Representative ExperimentalBasins, pp. 478–93. International Association ofScientific Hydrologists, Wallingford.

Schick, A.P. (1988) Hydrologic aspects of floods inextreme arid environments. In: Flood Geomorpho-logy (Eds V.R. Baker, R.C. Kochel & P.C. Patton), pp. 189–203. Wiley, New York.

ESD01 9/5/06 3:09 PM Page 418

REFERENCES 419

Schindler, C., Cuenod, Y., Eisenlohr, T., et al. (1993)The events of Randa, April 18th and May 9th, 1991 –an uncommon type of rockfall. Ecologae GeologicaeHelvetiae 86(3), 643–65.

Schlager, W. (2003) Benthic carbonate factories of thePhanerozoic. International Journal of Earth Science92, 445–64.

Schlyter, P., Jönsson, P., Nyberg, R., et al. (1993)Geomorphic process studies related to climate changein Kärkevagge, Norhern Sweden – status of research.Geografiska Annaler 75A(1–2), 55–60.

Schmidt, K.H. (1980) Eine neue Methode sur Ermitt-lung von Stufenrückwanderungsratendargestellt amBeispiel der Black Mesa Schichtstufe, Colorado Plateau,USA. Zeitschrift für Geomorphologie 24, 180–91.

Schmidt, K.H. (1989) The significance of scarp retreat forCenozoic landform evolution on the Colorado Plateau.Earth Surface Processes and Landforms 14, 93–105.

Schöne, B.R., Freyre Castro, A.D., Feibig, J., et al.(2004) Sea surface water temperatures over the period1884–1983 reconstructed from oxygen isotope ratiosof a bivalve mollusc shell (Arctica islandica, southernNorth Sea). Palaeogeography, Palaeoclimatology,Palaeoecology 212, 215–32.

Schrott, L., Hufschmidt, G., Hankammer, M., et al.(2003) Spatial distribution of sediment storage typesand quantification of valley fill deposits in an alpinebasin, Reintal, Bavarian Alps, Germany. Geomorpho-logy 55, 45–63.

Schumm, S.A. (1968) River Adjustment to AlteredHydrologic Regimen – Murrumbidgee River andPalaeochannels, Australia. Professional Paper 598, USGeological Survey, Denver, CO.

Schumm, S.A. (1969) River metamorphosis. Journal ofthe Hydraulic Division, American Society of CivilEngineers 95 (HY1), 255–73.

Schumm, S.A. & Chorley, R.J. (1966) Talus weatheringand scarp recession in the Colorado Plateaus. Zeitschriftfür Geomorphologie, Supplementband 10, 11–35.

Schumm, S.A. & Lichty, R.W. (1963) Channel wideningand flood-plain construction along Cimarron River in south-western Kansas. United States GeologicalSurvey Professional Paper 352D, 71–88.

Schumm, S.A. & Lichty, R.W. (1965) Time, space andcausality in geomorphology. American Journal ofScience 263, 110–19.

Schwartz, M. (1967) The Bruun theory of sea-level rise as a cause of shore erosion. Journal of Geology75, 76–92.

Schwarzer, K., Diesing, M., Larson, M., et al. (2003)Coastline evolution at different time scales: examplesfrom the Pomeranian Bight, southern Baltic sea. MarineGeology 194, 79–101.

Scoffin, T.P. (1970) The trapping and binding of subtidalcarbonate sediments by marine vegetation in Biminilagoon, Bahamas. Journal of Sedimentary Petrology40, 249–73.

Scoffin, T.P. (1987) An Introduction to CarbonateSediments and Rocks. Blackie, Glasgow.

Scoffin, T.P. (1993) The geological effects of hurricaneson coral reefs and the interpretation of storm deposits.Coral Reefs 12, 203–21.

Scoffin, T.P., Stearn, C.W., Boucher, D., et al. (1977)Calcium carbonate budget of a fringing reef on thewest coast of Barbados. Bulletin of Marine Science 30, 475–508.

Scott, D.B., Medioli, F.S., Schafer, C.T. (2001) Monitor-ing in coastal environments using Foraminifera andThecamoebian indicators. Cambridge University Press,192 pp.

Scott, K.M. (1988) Origins, Behaviour, and Sedimento-logy of Lahars and Lahar-runout Flows in the Toutle–Cowlitz River System. Professional Paper 1447A, USGeological Survey, Denver, CO, 74 pp.

Scott, K.M., Jana, R.J., de la Cruz, E.G., et al. (1996)Channel and sedimentation responses to large volumesof 1991 volcanic deposits on the east flank of MountPinatubo. In: Fire and Mud, Eruptions and Lahars ofMount Pinatubo, Philippines (Eds C.G. Newhall &R.S. Punongbayan), pp. 971–88. PHIVOLCS Press,Quezon City, and University of Washington Press,Seattle.

Scourse, J.D., Austin, W.E.N., Long, B.T., et al. (2002)Holocene evolution of seasonal stratification in theCeltic Sea: refined age model, mixing depths and for-aminiferal stratigraphy. Marine Geology 191, 119–45.

Scruton, P.C. (1960) Delta building and the deltaicsequence. In: Recent Sediments. North West Gulf of Mexico (Eds F.P. Shepard, F.B. Phleger & T.H. van Andel), pp. 82–102. American Association ofPetroleum Geologists, Tulsa, OK.

Sear, D.A., Newson, M.D. & Brookes, A. (1995)Sediment-related river maintenance: the role of fluvialgeomorphology. Earth Surface Processes and Land-forms 20, 629–47.

Sear, D.A., Lee, M.W.E., Oakey, R.J., Carling, P.A. &Collins, M.B. (2000) Coarse sediment tracing techno-logy in littoral and fluvial environments: a review. In: Tracers in Geomorphology (Ed. I.D.L. Foster), pp. 21–55. Wiley, Chichester.

Selley, R.C. (1994) Applied sedimentology. AcademicPress, London.

Semadeni-Davies, A. (2004) Urban water managementvs climate change: impacts on cold region waste waterinflows. Climatic Change 64, 103–26.

Semeniuk, V. (1981) Long-term erosion of the tidal flats,King Sound, north western Australia. Marine Geology43, 21–48.

Semeniuk, V. (1994) Predicting the effect of sea-level riseon mangroves in northwestern Australia. Journal ofCoastal Research 10, 1050–76.

Serrano-Belles, C. & Leharne, S. (1997) Assessing thepotential for lead release from road dusts and soils.Environmental Geochemistry and Health 19, 89–100.

ESD01 9/5/06 3:09 PM Page 419

420 REFERENCES

Shackleton, N.J., Berger, A. & Peltier, W.R. (1990) An alternative astronomical calibration of the lowerPleistocene timescale based on ODP Site 677. Trans-actions of Royal Society of Edinburgh: Earth Sciences81, 251–61.

Shakesby, R.A., Chafer, C.J., Doerr, S.H., et al. (2003)Fire severity, water repellency characteristics and hydro-geomorphological changes following the Christmas2001 Sydney forest fires. Australian Geographer 34,147–75.

Sharon, D. (1974) The spatial pattern of convective rain-fall in Sukumaland, Tanzania – a statistical analysis.Archiv für Meteorologie, Geophysik und Bioklimato-logie, Series B 22, 201–18.

Shaw, J. (1985) Beach morphodynamics of an AtlanticCoast embayment – Runkerry Strand, NorthernIreland. Irish Geography 18, 50–8.

Shaw, P.A. & Thomas, S.G. (1997) Pans, playas and saltlakes. In: Arid Zone Geomorphology: Process, Formand Change in Drylands, 2nd edn (Ed. D.S.G. Thomas),pp. 293–317. Wiley, Chichester.

Shaw, T.L. (1995) Environmental effects of estuary barrages. Proceedings of the Institute of Civil Engineers:Water, Maritime and Engineering 112, 48–59.

Sheppard, C.R.C. & Salm, R.V. (1988) Reef and coralcommunities of Oman, with a description of a newcoral species (Order Scleractinia, Genus Acanthastrea).Journal of Natural History 22, 263–79.

Sherman, D. & Bauer, B.B. (1993) Dynamics of beach-dune systems. Progress in Physical Geography17, 413–47.

Sherman, D.J., Jackson, D.W.T. & Namikas, S. (1998)Wind blown sand on beaches: an evaluation of models. Geomorphology 22, 113–33.

Sherman, D.J., Barron, K.M. & Ellis, J.T. (2002)Retention of beach sands by dams and debris basins in southern California. Journal of Coastal Research(Special Issue) 36, 662–74.

Shi, Z., Hamilton, L.J. & Wolanski, E. (2000) Near-bedcurrents and suspended sediment transport in saltmarshcanopies. Journal of Coastal Research 16, 909–14.

Shine, J.P., Ika, R. & Ford, T.E. (1998) Relationshipbetween oxygen consumption and sediment-waterfluxes of heavy metals in coastal sediments. Environ-mental Toxicology and Chemistry 17, 2325–7.

Short, A.D. (1999) Handbook of Beach and ShorefaceMorphodynamics. Wiley, Chichester, 379 pp.

Siebert, L. (1984) Large volcanic debris avalanches:characteristics of source areas, deposits and associatederuptions. Journal of Volcanology and GeothermalResearch 22(304), 163–97.

Silvester, R. (1976) Headland defense of coasts. CoastalEngineering 1, 1394–406.

Simmons, I.G. & Innes, J.B. (1996) Prehistoric charcoalin peat profiles at North Gill, North YorkshireMoores, England. Journal of Archaeological Science23, 193–7.

Singh, B.P., Pawar, J.S. & Karlupia, S.K. (2004) Densemineral data from the north-western Himalayan fore-land sedimentary rocks and recent river sediments:evaluation of the hinterland. Journal of Asian EarthSciences 23, 25–35.

Sivan, D., Wdowinski, S., Lambeck, K., et al. (2001)Holocene sea-level changes along the Mediterraneancoast of Israel, based on archaeological observationsand numerical model. Palaeogeography, Palaeo-climatology, Palaeoecology 167, 101–17.

Slaymaker, O., Souch, C., Menounos, B., et al. (2003)Advances in Holocene mountain geomorphologyinspired by sediment budget methodology. Geomor-phology 55, 305–16.

Sly, P.G. (1978) Sedimentary processes in lakes. In:Lakes: Chemistry, Geology, Physics (Ed. A. Lerman),pp. 65–89. Springer-Verlag, Berlin.

Small, D.L., Payonk, P.M. & Jarrett, J.T. (1997) Bene-ficial use of dredged material in nearshore placementareas in North Carolina. In: Proceedings: InternationalWorkshop on Dredged Material Beneficial Uses (Ed.M.C. Landin), p. 152. Unnumbered U.S. Army Corpsof Engineers Document, Waterways ExperimentStation, Vicksburg, Mississippi.

Smith, B. & Warke, P. (1997) Controls and uncertaintyin the weathering environment. In: Arid Zone Geo-morphology: Process, Form and Change in Drylands,2nd edn (Ed. D.S.G. Thomas), pp. 41–54. Wiley,Chichester.

Smith, B.A. & Marchand, M. (2001) Coastal erosionand catchment management. Coastline 10(3), 3–6.

Smith, K. (2001) Environmental Hazards – AssessingRisk and Reducing Disaster, 3rd edn. Routledge,London.

Smith, N.D. (1974) Sedimentology and bar formation in the Upper Kicking Horse River, a braided outwashstream. Journal of Geology 82, 205–24.

Smith, T.J. (1992) Forest Structure. In: TropicalMangrove Ecosystems (Eds A.I. Robertson & D.M.Alongi), pp. 101–36. Coastal and Estuarine Studies 41,American Geophysical Union, Washington.

Smith, T.J., Robblee, M.B., Wanless, H.R., et al.(1994) Mangroves, hurricanes, and lighting strikes.BioScience 44, 256–62.

Smithers, S. & Larcombe, P. (2003) Late Holocene initiation and growth of a nearshore turbid-zone coral reef: Paluma Shoals, central Great Barrier Reef,Australia. Coral Reefs 22, 499–505.

Smol, J.P. (2002) Pollution of Lakes and Rivers: aPaleoenvironmental Perspective. Arnold, London.

Snedden, J.W. & Nummedal, D. (1991) Origin andgeometry of storm-deposited sand beds in modernssediments of the Texas continental shelf. In: Shelf Sandand Sandstone Bodies (Eds D.J.P. Swift, G.F. Oertel,R.W. Tillman & J.A. Thorne), pp. 283–308. SpecialPublication 14, International Association of Sedi-mentologists. Blackwell Scientific, Oxford.

ESD01 9/5/06 3:09 PM Page 420

REFERENCES 421

Södergren, A. (Ed.) (1992) Bleached Pulp Mill Effluents.Composition, Fate and Effects in the Baltic Sea.Report 4047, Swedish Environmental ProtectionAgency, Stockholm, 150 pp.

Södergren, A., Jonsson, P., Bengtsson, B.-E., et al. (1988)Biological Effects of Bleached Pulp Mill Effluents.Report 3558, Swedish Environmental Protection,Stockholm, 134 pp.

Soldati, M., Corsini, A. & Pasuto, A. (2004) Landslidesand climate change in the Italian Dolomites since thelate glacial. Catena 55, 141–61.

Solomon, S.M. & Forbes, D.L. (1999) Coastal hazardsand associated management issues on South PacificIslands. Ocean & Coastal Management 42, 523–54.

Sommerfield, C.K., Nittrouer, C.A. & DeMaster, D.J.(1996) Carbon-isotope systematics on the Amazonshelf. Geomarine Letters 16, 17–23.

Soto-Jiménez, M.F. & Páez-Osuna, F. (2001) Distribu-tion and normalization of heavy metal concentrationsin mangrove and lagoonal sediments from MazatlánHarbor (SE Gulf of California). Estuarine, Coastaland Shelf Science 53, 259–74.

Soulsby, R. (1997) Dynamics of Marine Sands. ThomasTelford, London.

Soulsby, R.L. (1983) The bottom boundary layer of shelf seas. In: Physical Oceanography of Coastal andShelf Seas (Ed. B. Johns), pp. 189–266. Elsevier,Amsterdam.

Spalding, M.D. & Grenfell, A.M. (1997) New estimatesof global and regional coral reef areas. Coral Reefs16, 225–30.

Spalding, M.D., Blasco, F., Field, C.D. (1997) WorldMangrove Atlas. International Society for MangroveEcosystems, Okinawa, Japan.

Spedding, N. (2000) Hydrological controls on sedimenttransport pathways: implications or debris-coveredglaciers. In: Debris-covered Glaciers, Workshop heldat Seattle, September, pp. 133–42. IAHS Publication264, International Association of Hydrological SciencesPress, Wallingford.

Spencer, T. (1995) Potentialities, uncertainties and complexities in the response of coral reefs to futuresea-level rise. Earth Surface Processes and Landforms20, 49–64.

Speth, J.G. (1994) Towards an Effective and Opera-tional International Convention on Desertification.International Negotiating Committee, InternationalConvention on Desertification, United Nations, NewYork.

Stanley, D.J. & Warne, A.G. (1998) Nile delta in itsdestruction phase. Journal of Coastal Research 14,794–825.

Starkel, L. (1983) The reflection of hydrologic changes in the fluvial environment of the temperate zone during the last 15,000 years. In: Background toPalaeohydrology: a Perspective (Ed. K.J. Gregory),pp. 213–35. Wiley, Chichester.

Staub, J.R., Among, H.L. & Gastaldo, R.A. (2000)Seasonal sediment transport and deposition in theRajang River delta, Sarawak, East Malaysia. Sedi-mentary Geology 133, 249–64.

Sternberg, R.W. & Newell, A.R.M. (1999) Continentalshelf sedimentology: scales of investigation definefuture research opportunities. Journal of Sea Research41, 55–71.

Stevenson, A.G. (2001) Metal concentrations in marinesediments around Scotland: a baseline for environ-mental studies. Continental Shelf Research 21,879–97.

Stigliani, W.M. (Ed.) (1991) Chemical Time Bombs:Definition, Concepts, and Examples. ExecutiveReport 16, International Institute of Applied SystemsAnalysis, Laxenburg, Austria.

Stoddart, D.R. (1980) Mangroves as successional stages, inner reefs of the Great Barrier Reef. Journal ofBiogeography 7, 269–84.

Stoddart, D.R. (1990) Coral reefs and islands and pre-dicted sea-level rise. Progress in Physical Geography14, 521–36.

Stoddart, D.R., Reed, D.G. & French, J.R. (1989)Understanding salt-marsh accretion, Scolt Head Island,Norfolk, England. Estuaries 12, 228–36.

Stoermer, E.F. & Smol, J.P. (1999) The Diatoms:Applications for the Environmental and Earth Sciences.Cambridge University Press, Cambridge, 469 pp.

Stokes, G.G. (1851) Collected Papers, Vol. III. Cam-bridge Transactions Vol. IX (see e.g. Lamb, H., 1945.Hydrodynamics. Gover Publications, New York, 450 pp.)

Stokes, W.L. (1968) Multiple parallel truncation bedding planes – a feature of wind-deposited sand-stone formations. Journal of Sedimentary Petrography38, 510–15.

Stone, M. & Marsalek, J. (1996) Trace metal com-position and speciation in street sediment: Sault Ste,Marie, Canada. Water, Air and Soil Pollution 87,149–69.

Stow, D.A.V. (1986) Deep clastic seas. In: SedimentaryEnvironments and Facies, 2nd edn (Ed. H.G. Reading),pp. 399–470. Blackwell Science, Oxford.

Stride, A.H. (Ed.) (1982) Offshore Tidal Sands.Chapman and Hall, London, pp. 95–125.

Stum, W. (1992) Chemistry of the Solid–Water InterfaceWiley Interscience, New York.

Sturm, M. (1975) Depositional and erosional sedi-mentary features in a turbidity current controlledbasin (Lake brienz). IX International Congress ofSedimentology, Nice, Theme 5, Vol. 5/2.

Sturm, M. & Matter, A. (1978) Turbidites and varves inLake Brienz (Switzerland): deposition of clastic detritusby density currents. In: Modern and Ancient LakeSediments (Eds A. Matter & M.E. Tucker), pp. 147–68. Special Publication 2, International Association ofSedimentologists. Blackwell Scientific, Oxford.

ESD01 9/5/06 3:09 PM Page 421

422 REFERENCES

Sullivan, M.J. (1999) Applied diatom studies in estuariesand shallow coastal waters. In: The Diatoms: Applica-tions for the Environmental and Earth Sciences (EdsE.F. Stoermer & J.P. Smol), pp. 334–51. CambridgeUniversity Press, Cambridge.

Summerfield, M.A. (1991) Global Geomorphology.Longman, New York, 537 pp.

Summerfield, M.A. & Hulton, N.J. (1994) Natural controls of fluvial denudation rates in major worlddrainage basins. Journal of Geophysical Research99(B7), 13871–83.

Surian, N. & Rinaldi, M. (2003) Morphological responseto river engineering and management in alluvial channels in Italy. Geomorphology 50, 307–26.

Sutherland, R.A. (2003) Lead in grain size fractions of road-deposited sediment. Environmental Pollution121, 229–7.

Swennen, R., Van Keer, I. & De Vox, W. (1994) Heavymetal contamination in overbank sediments of theGeul river (East Belgium): its relation to former Pb-Zn mining activities. Environmental Geology 24,12–21.

Swiadeck, J.W. (1997) The impacts of Hurricane Andrewon mangrove coasts in southern Florida: a review.Journal of Coastal Research 13, 242–5.

Swift, D.J.P., Han, G. & Vincent, C.E. (1986) Fluid processes and sea-floor responses on a modern storm-dominated shelf: middle-Atlantic shelf of North America. Part 1. The storm current regime. In:Shelf Sands and Sandstones (Eds R.J. Knight & J.R. McLean), pp. 99–119. Memoir 11, CanadianSociety of Petroleum Geologists, Calgary.

Swinchatt, J.P. (1965) Significance of constituent com-position, texture and skeletal breakdown in someRecent carbonate sediments. Journal of SedimentaryPetrology 35, 71–90.

Szmant, A.M. & Forrester, A. (1996) Water column and sediment nitrogen and phosphorous distributionpatterns in the Florida Keys, USA. Coral Reefs 15,21–42.

Taylor, K.G., Boyd, N.A. & Boult, S. (2003) Sediments,porewaters and diagenesis in an urban water body,Salford, UK: impacts of remediation. HydrologicalProcesses 17, 2049–61.

Te, F.E. (1997) Turbidity and its effect on corals: amodel using the extinction coefficient (k) of photo-sythetic active radiation (PAR). Proceedings of EighthInternational Coral Reef Symposium, Panama 2,1899–904.

Teal, J.M., Farrington, J.W., Burns, K.A., et al. (1992)The west Falmouth oil spill after 20 years: fate of fuel oil compounds and effects on animals. MarinePollution Bulletin 24, 607–14.

Teller, J.T. (2001) Formation of large beaches in an areaof rapid differential isostatic rebound: the three-outletcontrol of Lake Agassiz. Quaternary Science Reviews20, 1649–59.

Testa, V. & Bosence, E.J. (1999) Physical and biologicalcontrols on the formation of carbonate and silici-clastic bedforms on the north-east Brazilian shelf.Sedimentology 46, 279–301.

Thampanya, U., Vermaat, J.E. & Terrados, J. (2002)The effects of increasing sediment accretion on theseedlings of three common Thai mangrove species.Aquatic Biology 74, 315–25.

The Joint Nature Conservation Committee (2004) TheIrish Sea Pilot Final Report. 176 pp. (available atwww.jncc.gov.uk).

The Times Atlas of the World (1983) ComprehensiveEdition, 6th edn. John Bartholomew and Times BooksLimited, London.

Thieler, E.R., Pilkey, O.H., Young, R.S., et al. (2000)The use of mathematical models to predict beachbehavior for U.S. coastal engineering: a critical review.Journal of Coastal Research 16, 48–70.

Thom, B.G. (1967) Mangrove ecology and deltaic geo-morphology: Tabasco, Mexico. Journal of Ecology55, 301–43.

Thom, B.G. (1982) Mangrove ecology: a geomorpho-logical perspective. In: Mangrove Ecosystems inAustralia, Structure, Function and Management(Ed. B.F. Clough), pp. 3–17. Australian NationalUniversity Press, Canberra.

Thom, B.G. (1984) Transgressive and regressive strati-graphies of coastal sand barriers in southeast Australia.Marine Geology 56, 137–58.

Thomas, D.S.G. (1989) Reconstructing ancient aridenvironments. In: Arid Zone Geomorphology: Process,Form and Change in Drylands, 2nd edn (Ed. D.S.G.Thomas), pp. 311–34. Wiley, Chichester.

Thomas, D.S.G. (1997a) Arid environments. In: AridZone Geomorphology: Process, Form and Change inDrylands, 2nd edn (Ed. D.S.G. Thomas), pp. 3–12.Wiley, Chichester.

Thomas, D.S.G. (1997b) Sand seas and aeolian bedforms.In: Arid Zone Geomorphology: Process, Form andChange in Drylands, 2nd edn (Ed. D.S.G. Thomas),pp. 273–412. Wiley, Chichester.

Thomas, D.S.G. (1997c) Science and the desertificationdebate. Journal of Arid Environments 37, 599–608.

Thomas, D.S.G. & Middleton, N.J. (1994) Deserti-fication: Exploding the Myth. Wiley. Chichester, 1194 pp.

Thomas, M.G. (1998) Late Quaternary landscape instability in the humid and sub-humid tropics. In:Palaeohydrology and Environmental Change (Eds G. Benito, V.R. Baker & K.J. Gregory), pp. 247–58.Wiley, Chichester.

Thomas, R.L., Jacquet, J.M., Kemp, A.L.W., et al.(1976) Surficial sediments in Lake Erie. Journal of theFisheries Research Board of Canada 33, 385–403.

Thomas, S. & Ridd, P.V. (2004) Review of methods to measure short time scale sediment accumulation.Marine Geology 207, 95–114.

ESD01 9/5/06 3:09 PM Page 422

REFERENCES 423

Thomas, S., Ridd, P.V. & Day, G. (2003) Turbidityregimes over fringing coral reefs near a mining site atLihir Island, Papua New Guinea. Marine PollutionBulletin 46, 1006–14.

Thompson, T.A. & Baedke, S.J. (1995) Beach-ridgedevelopment in Lake Michigan: shoreline behavior inresponse to quasi-periodic lake-level events. MarineGeology 129, 163–74.

Thorne, P.D. & Hanes, D.M. (2002) A review of acoustic measurement of small-scale sediment pro-cesses. Continental Shelf Research 22, 603–32.

Thorne, R.D., Hey, R.D. & Newson, M.D. (Eds) (1997) Applied Fluvial Geomorphology for RiverEngineering. Wiley, Chichester.

Thornes, J.B. (1974) The rain in Spain. GeogographicalMagazine 47, 337–43.

Thornes, J.B. (1977) Channel changes in ephemeralstreams: observations, problems and models. In: RiverChannel Changes (Ed. R.C. Gregory), pp. 353–67.Wiley, Chichester.

Thornthwaite, C.W. (1948) An approach towards arational classification of climate. Geographical Review38, 55–94.

Thornton, I., Watt, J.M., Davies, D.J.A., et al. (1994)Lead contamination of UK dusts and soils and implications for childhood exposure – an overview of the work of the Environmental GeochemistryResearch Group, Imperial College, London, England,1981–1992. Environmental Geochemistry and Health16, 113–22.

Thurman, E.M. (1985) Organic Geochemistry of NaturalWaters. Martinus Nijhoff/Dr W. Junk Publishers,Dordrecht, 487 pp.

Tomascik, T. & Sander, F. (1985) Effects of eutro-phication on reef-building corals I. growth rate of the reef-building coral Montastrea annularis. MarineBiology 87, 143–55.

Tomascik, T. & Sander, F. (1987) Effects of eutrophica-tion on reef-building corals. III. Reproduction of thereef-building coral Porites porites. Marine Biology 94, 77–94.

Tomlinson, P.B. (1986) The Botany of Mangroves.Cambridge University Press, Cambridge.

Topçu, D.H. & Brockmann, U. (2001) Synthesis and NewConception of North Sea Research (SYCON) WorkingGroup 5: Fluxes of Matter. Zentrum für Meeres – undKlimaforschung der Universität, Hamburg, 167 pp.

Törnqvist, T.E. (1993) Holocene alteration of meander-ing and anastomising fluvial systems in the Rhine–Meuse delta (central Netherlands) controlled by sealevel rise and subsoil erodability. Journal of SedimentaryPetrology 63, 683–93.

Törnqvist, T.E. (1994) Middle and late Holocene avulsion history of the River Rhine (Rhine–MeuseDelta, Netherlands). Geology 22, 711–14.

Toscano, M.A. & Macintyre, I.G. (2003) Correctedwestern Atlantic sea-level curve for the last 11,000 years

based on calibrated 14C dates from Acropora palmataframework and intertidal mangrove peats. Coral Reefs22, 257–70.

Trauth, M.H., Alfonso, R.A. & Haselton, K.R. (2000)Climate change and mass movement in the NWArgentine Andes. Earth and Planetary Science Letters179, 243–56.

Trefry, J.H. & Presley, B.J. (1976) Heavy metal trans-port from the Mississippi River to the Gulf of Mexico.In: Marine Pollution Transfer (Eds H.L. Windom & R.A. Duce), pp. 39–76. Health Publications,Lexington.

Trenhaile, A.S. (1997) Coastal Dynamics and Landforms.Clarendon Press, Oxford.

Trichet, J., Defarge, C., Tribble, J., et al. (2001)Christmas Island lagoonal lakes, models for the deposition of carbonate–evaporite–organic laminatedsediments. Sedimentary Geology 140, 177–89.

Trimmer, M., Petersen, J., Sivyer, D.B., et al. (2005)Impacts of long-term benthic trawl disturbance onsediment sorting and biogeochemistry in the southernN. Sea. Marine Ecology Progress Series 298, 79–94.

Trustrum, N.A., Gomez, B., Page, M.J., et al. (1999)Sediment production, storage and output: the relativerole of large magnitude events in steepland catchments.Zeitschrift für Geomorphologie (N.F. supplement)115, 71–86.

Tucker, C.J., Dregne, H.E. & Newcomb, W.W. (1991)Expansion and contraction of the Sahara Desert from1980 to 1990. Science 253, 299.

Tucker, M.E. & Wright, V.P. (1990) Carbonate Sedi-mentology. Blackwells, Oxford.

Tudhope, A.W. & Scoffin, T.P. (1984) The effects of Callianassa bioturbation on the preservation of carbonate grains in Davies reef lagoon, Great BarrierReef, Australia. Journal of Sedimentary Petrology 54,1091–6.

Udvardy, M.D.F. (1981) The riddle of dispersal: dis-persal theories and how they affect vicariance bio-geography. In: Vicariance Biogeography: a Critique(Eds G. Nelson & D.E. Rosen), pp. 6–29. ColumbiaUniversity Press, New York.

UN (1977) Status of desertification in the hot aridregions, climatic aridity index map and experimentalworld scheme of aridity and drought probability, at a scale of 1:25,000,000. Explanatory note. UNConference on Desertification A/CONF. 74/31, NewYork.

Underwood, G.J.C. & Smith, D.J. (1998) In-situ meas-urements of exopolymer production by intertidalepipelic diatom-dominated biofilms. In: SedimentaryProcesses in the Intertidal Zone (Eds K.S. Black, D.H. Paterson & A. Cramp), pp. 125–34. SpecialPublication 139, Geological Society Publishing House,Bath.

UNEP (1992) World Atlas of Desertification. EdwardArnold, Sevenoaks, 69 pp.

ESD01 9/5/06 3:09 PM Page 423

424 REFERENCES

UNESCO (1979) Map of the World Distribution of AridRegions. Technical Note 4, Man and the BiosphereProgramme, UNESCO, New York.

United Nations (2003) Water and Sanitation in theWorld’s Cities: Local Action for Global Goals. Press Release HAB/184, 19 March, UN-HABITAT,Nairobi.

USACE (2003) Shore Protection Manual. United StatesArmy Corps of Engineers, Vicksburg, Mississippi. Avail-able at: http://bigfoot.wes.army.mil/cem026.html

USACE/EPA (1992) Evaluating Environmental Effectsof Dredged Material Management Alternatives – aTechnical Framework. EPA842-B-92-008, Environ-mental Protection Agency, Office of Marine andEstuarine Protection, Washington, DC.

Van Dalfsen, J.A., Essink, K., Toxvig Madsen, H., et al.(2000) Differential response of macrozoobenthos tomarine sand extraction in the North Sea and theWestern Mediterranean. ICES Journal of MarineScience 57, 1439–45.

Van den Bergh, G.D., Boer, W., de Hass, H., et al. (2003)Shallow marine tsunami deposits in Teluk Banten(NW Java, Indonesia), generated by the 1883 Krakataueruption. Marine Geology 197, 13–34.

Van Walree, P.A., Tjegowski, J., Labanc, C., et al.(2005) Acoustic seafloor discrimination with echoshape parameters: a comparison with the groundtruth. Continental Shelf Research 25, 2273–93.

Van Woesik, R., De Vantier, L.M. & Glazebrook, J.S.(1995) Effects of Cyclone ‘Joy’ on nearshore coralcommunities of the Great Barrier Reef. Marine EcologyProgress Series 128, 261–70.

Vaughan, N.D., Johnson, T.C., Mearns, D.L., et al.(1987) The impact of Hurricane Diana on the North Carolina continental shelf. Marine Geology 76,169–76.

Veldkamp, A. & Van Dijke, J.J. (1998) Modelling long-term erosion and sedimentation processes in fluvialsystems: a case study for the Allier/Loire system In:Palaeohydrology and Environmental Change (Eds G. Benito, V.R. Baker & K.J. Gregory), pp. 53–66.Wiley, Chichester.

Verhagen, H. (1992) Method for artificial beach renourishment, Proceedings of the 23rd InternationalConference on Coastal Engineering, Venice, Italy, pp. 2474–85.

Viklander, M. (1998) Particle size distribution and metalcontent in street sediments. Journal of EnvironmentalEngineering 124, 761–66.

Viles, H. & Spencer, T. (1995) Coastal Problems:Geomorphology, Ecology and Society at the Coast.Edward Arnold, London.

Viseras, C., Calvache, M.L., Soria, J.M., et al. (2003)Differential features of alluvial fans controlled by tectonic or eustatic accommodation space, examplesfrom the Betic Cordillera, Spain. Geomorphology50, 181–20.

Vivian, C.M.G. & Murray, L.A. (2001) Pollution, solids.In: Encyclopaedia of Ocean Sciences (Eds J. Steele, S. Thorpe & K.K. Turekian), pp. 2236–41. AcademicPress, London.

Vollenweider, R.A. (1968) The Scientific Basis of Lake Eutrophication, with Particular Reference toPhosphorus and Nitrogen as Eutrophication Factors.Technical Report DAS/DSI/68.27, Organization ofEconomic Cooperation and Development, Paris, 159 pp.

Vollenweider, R.A. (1976) Advances in defining criticalloading levels for phosphorus in lake eutrophication.Istituto Italiano di Idrobiologia 33, 53–83.

Walder, J.S. & Driedger, C.L. (1995) Frequent outburstfloods from South Tahoma Glacier, Mount Rainier,USA: relations to debris flows, meteorological originand implications for subglacial hydrology. Journal ofGlaciology 41(137), 1–10.

Walker, R.G. & James, N.P. (Eds) (1992) Facies Models:Response to Sea-level Change. Geological Associationof Canada, St John’s, Newfoundland.

Walker, W.J., McNutt, R.P. & Maslanka, C.K. (1999)The potential contribution of urban runoff to surfacesediments of the Passaic River: sources and chemicalcharacteristics. Chemosphere 38, 363–77.

Wallin, M., Håkanson, L. & Persson, J. (1992) Loadmodels for nutrients in coastal areas, especially fromfish farms (in Swedish with English summary).Nordiska ministerrådet, Copenhagen 502, 207 pp.

Walling, D.E. (1987) Rainfall, runoff and erosion of the land: a global view. In: Energetics of PhysicalEnvironment (Ed. K.J. Gregory), pp. 89–117. Wiley,Chichester.

Walling, D.E. (1990) Linking the field to the river: sediment delivery from agricultural land. In: SoilErosion on Agricultural Land (Eds J. Boardman,I.D.L. Foster & J.A. Dearing), pp. 129–52. Wiley,Chichester.

Walling, D.E. (1999) Linking land use, erosion and sediment yields in river basins. Hydrobiologia 410,223–40.

Walling, D.E. & Fang, D. (2003) Recent trends in thesuspended sediment loads of the world’s rivers.Global and Planetary Change 39, 111–26.

Walling, D.E. & Kleo, A.H.A. (1979) Sediment yields of rivers in areas of low precipitation: a global view.In: The Hydrology of Areas of Low Precipitation,Proceedings of the Canberra Symposium, December,pp. 479–93. IAHS Publication 128, InternationalAssociation of Hydrological Sciences Press, Wallingford.

Walling, D.E. & Webb, B.W. (1983) Patterns of sedi-ment yield. In: Background to Palaeohydrology: aPerspective (Ed. K.J. Gregory), pp. 69–100. Wiley,Chichester.

Walling, D.E., Peart, M.R., Oldfield, F., et al. (1979)Suspended sediment sources identified by magneticmeasurements. Nature 281, 110–13.

ESD01 9/5/06 3:09 PM Page 424

REFERENCES 425

Walling, D.E., Woodward, J.C. & Nicholas, A.P. (1993)A multi-parameter approach to fingerprinting sus-pended sediment sources. In: Erosion and SedimentTransport Monitoring Programmes (Eds J. Bogen,D.E. Walling & T. Day), pp. 329–38. IAHS Publica-tion 215, International Association of HydrologicalSciences Press, Wallingford.

Walling, D.E., Owens, P.N. & Leeks, G.J.L. (1999)Fingerprinting suspended sediment sources in thecatchment of the River Ouse, Yorkshire, UK. Hydro-logical Processes 13, 955–75.

Walling, D.E., Russell, M.A., Hodgkinson, R.A., et al.(2002) Establishing sediment budgets for two smalllowland agricultural catchments in the UK. Catena47, 323–53.

Walling, D.E., Owens, P.N., Carter, J., et al. (2003)Storage of sediment-associated nutrients and con-taminants in river channels and floodplain systems.Applied Geochemistry 18, 195–220.

Warburton, J. (1990) An alpine proglacial fluvial sedi-ment budget. Geografiska Annaler 72A(3–4), 261–72.

Warren, A. (1979) Aeolian processes. In: Process inGeomorphology (Eds C. Embleton & J. Thornes), pp. 325–551. Edward Arnold, Sevenoaks.

Warren, N., Allan, I.J., Carter, J.E., et al. (2003)Pesticides and other micro-organic contaminants in freshwater sedimentary environments – a review.Applied Geochemistry 18, 159–94.

Wasson, R.J. & Hyde, R. (1983) Factors determiningdesert dune type. Nature 304, 337–9.

Watling, L. & Norse, E.A. (1998) Disturbance of theseabed by mobile fishing gear: a comparison to forestclearcutting. Conservation Biology 12, 1180–97.

Watson, A. & Nash, D.J. (1997) Desert crusts and varnishes. In: Arid Zone Geomorphology: Process,Form and Change in Drylands, 2nd edn (Ed. D.S.G.Thomas), pp. 69–107. Wiley, Chichester.

Watts, C.D., Naden, P.S., Cooper, D.M., et al. (2003)Application of a regional procedure to assess the riskto fish from high sediment concentrations. The Scienceof the Total Environment 314–16, 551–65.

Webb, R.H., Pringle, P.T., Reneau, S.L., et al. (1988)Monument Creek debris flow, 1984 – implications forformation of rapids on the Colorado River in GrandCanyon National Park. Geology 6, 50–4.

Wei, C. & Morrison, G.M. (1994a) Platinum in roaddusts and urban river sediments. The Science of theTotal Environment 147, 169–74.

Wei, C. & Morrison, G.M. (1994b) Platinum analysisand speciation in urban gullypots. Analytica ChimicaActa 284, 587–92.

Wells, S.G. & Harvey, A.M. (1987) Sedimentalogic and geomorphic variations in storm-generated alluvialfans, Howgill Fells, northwest England. GeologicalSociety of America Bulletin 98, 182–98.

Wenban-Smith, F. (2002) Palaeolithic and MesolithicArchaeology on the Sea Bed: Marine Aggregate

Dredging and the Historic Environment. Report tothe British Marine Aggregate Producers Association,london, 18 pp. (available from www.bmapa.org.uk).

Wenning, R.J., Batley, G.R., Ingersoll, C.G., et al. (Eds)(2005) Use of sediment quality guidelines & relatedtools for the assessment of contaminated sediments.Society of Environmental Toxicology and Chemistry,816 pp.

Werritty, A. & Leys, K.F. (2001) The sensitivity ofScottish rivers and upland valley floors to recent environmental change. Catena 42, 251–73.

Westaway, R., Maddy, D. & Bridgland, D. (2002) Flowin the lower continental crust as a mechanism for theQuaternary uplift of south-east England: constraintsfrom the Thames terrace record. Quaternary ScienceReviews 21, 559–603.

Wetzel, D.L. & Van Vleet, E.S. (2003) Persistence ofpetroleum hydrocarbon contamination in sedimentsof the canals of Venice, Italy: 1995 and 1998. MarinePollution Bulletin 46, 1015–23.

Wetzel, R.G. (2001) Limnology and Lake Ecosystems,3rd edn. Academic Press, San Diego, 1006 pp.

Whipple, K.X. & Trayler, C.R. (1996) Tectonic controlof fan size: the importance of spatially variable sub-sidence rates. Basin Research 8, 351–66.

White, S., García-Ruiz, J.M., Martí, C., et al. (1997) The1996 Biescas campsite disaster in the Central SpanishPyrenees, and its temporal and spatial context. Hydro-logical Processes 11, 1797–812.

Whiteley, J.D. & Murray, F. (2003) Anthropogenic platinum group element (Pt, Pd and Rh) concentra-tions in road dusts and roadside soils from Perth,Western Australia. The Science of the Total Environ-ment 317, 121–35.

Whitney, M.I. & Dietrich, R.V. (1973) Ventifact sculpture by wind-blown dust. Geological Society ofAmerica Bulletin 84, 2561–82.

Whitten, D.G.A. & Brooks, J.R.V. (1972) The PenguinDictionary of Geology. Penguin Books, London, 514 pp.

Whittow, J. (1980) Landslides and avalanches – ava-lanches. In Disasters: the Anatomy of EnvironmentalHazards, pp. 163–70. Penguin, Harmondsworth.

Wicklund, A. (1990) Metabolism of cadmium and zincin fish. Thesis, Uppsala University, Sweden.

Wiedermann, H.V. (1972) Shell deposits and shellpreservation in Quaternary and Tertiary estuarinesediments in Georgia, USA. Sedimentary Geology7: 103–25.

Wiggs, G.F.S., Livingstone, I., Thomas, D.S.G., et al.(1994) The effect of vegetation removal on air flow structure and dune mobility in the southwestKalahari. Land Degradation and Rehabilitation 5,13–24.

Wilkinson, C.R. (1996) Global change and coral reefs:impacts on reefs, economies and human cultures.Global Change Biology 2, 547–58.

ESD01 9/5/06 3:09 PM Page 425

426 REFERENCES

Willets, B.B. (1983) Transport by wind of granularmaterials of different grain shapes and densities.Sedimentology 30, 669–79.

Willets, B.B. & Rice, M.A. (1985) Wind tunnel tracerexperiments using dyed sand. In: Proceedings of Inter-national Workshop on the Physics of Blown Sand, pp. 225–42. Memoir 8, Department of TheoreticalStatistics, Aarhus University, Denmark.

Williams, M.A.J. & Balling, R.C. (1995) Interactions ofDesertification and Climate. Edward Arnold, London,270 pp.

Williams, M.A.J. (1994) Cenozoic climatic changes in deserts: a synthesis. In: Geomorphology of DesertEnvironments (Eds A.D. Abrahams & A.J. Parsons),pp. 644–70. Chapman and Hall, London.

Wilson, I.G. (1973) Ergs. Sedimentary Geology 10,77–106.

Wilson, P. & Braley, S.M. (1997) Development and agestructure of Holocene coastal sand dunes at HornHead, near Dunfanaghy, Co. Donegal, Ireland. TheHolocene 7, 187–97.

Wilson, P., Orford, J.D., Knight, J., et al. (2001) Late-Holocene (post-4000 years BP) coastal dune develop-ment in Northumberland, northeast England. TheHolocene 11, 215–29.

Winkler, E.M. & Wilheim, E.J. (1970) Saltburst byhydration pressures in architectural stone in urbanatmosphere. Geological Society of America Bulletin81, 567–72.

Winkler, E.M. (1977) Insolation of rock and stone, a hotitem. Geology 5, 188–9.

Wisniak, J. & Garces, I. (2001) The rise and fall of thesalitre (sodium nitrate) industry. Indian Journal ofChemical Technology 8, 427–38.

Wolanski, E. (1995) Transport of sediment in mangroveswamps. Hydrobiologia 295, 31–42.

Wolanski, E. & Chappell, J. (1996) The response oftropical Australian estuaries to a sea level rise. Journalof Marine Systems 7, 267–79.

Wolanski, E., Jones, M. & Bunt, J.S. (1980) Hydro-dynamics of a tidal creek-mangrove swamp system.Australian Journal of Marine and Freshwater Research31, 431–50.

Wolanski, E., Mazda, Y. & Ridd, P. (1992) MangroveHydrodynamics. In: Tropical Mangrove Ecosystems(Eds A.I. Robertson & D.M. Alongi), pp. 43–62.Coastal and Estuarine Studies 41, American Geo-physical Union, Washington.

Wolanski, E., Gibbs, R.J., Spagnol, S., et al. (1998)Inorganic sediment budget in the mangrove-fringedFly River Delta, Papua New Guinea. Mangroves andSalt Marshes 2, 85–98.

Wolfenden, P.J. & Lewin, J. (1977) Distribution of metalpollutants in floodplain sediments. Catena 4, 309–17.

Wolman, M.G. & Gerson, R. (1978) Relative scales oftime and effectiveness of climate in watershed geomor-phology. Earth Surface Processes 3, 189–208.

Wood, P.A. (1978) Fine sediment mineralogy of sourcerocks and suspended sediments, Rother catchment,West Sussex. Earth Surface Processes 3, 255–63.

Woodhouse, W.W. (1978) Dune Building and Stabil-ization with Vegetation. Special Report 3, CoastalEngineering Research Center, U.S. Army CorpsEngineers, Springfield, VA.

Woodroffe, C. (1992) Mangrove sediments and geo-morphology. In: Tropical Mangrove Ecosystems (EdsA.I. Robertson & D.M. Alongi), pp. 7–41. Coastaland Estuarine Studies 41, American Geophysical Union,Washington.

Woodroffe, C.D. (1983) Development of mangrove forestsfrom a geological perspective. In: Tasks for VegetationScience 8 (Ed. H.J. Teas), pp. 1–17. Dr W. JunkPublishers, The Hague.

Woodroffe, C.D. (1990) The impact of sea-level rise onmangrove shorelines. Progress in Physical Geography14, 483–520.

Woodroffe, C.D. (2003) Coasts: Form, Process andEvolution. Cambridge University Press, Cambridge.

Woodroffe, C.D. & Grime, D. (1999) Storm impact and evolution of a mangrove-fringed chenier plain,Shoal Bay, Darwin, Australia. Marine Geology 159,303–21.

Woodroffe, C.D. & Grindrod, J. (1991) Mangrove biogeography: the role of Quaternary environmentaland sea-level change. Journal of Biogeography 18,479–92.

Woodroffe, C.D., Bardsley, K.N., Ward, P.J., et al.(1988) Production of mangrove litter in a macro-tidal embayment, Darwin Harbour, N.T., Australia.Estuarine, Coastal and Shelf Science 26, 581–98.

Woodward, J.C. (1995) Patterns of erosion and sus-pended sediment yield in Mediterranean river basins.In: Sediment and Water Quality in River Catchments(Eds I.D.L. Foster, A.W. Gurnell & B.W. Webb), pp. 365–89. Wiley. Chichester.

Woodward, J.C., Lewin, J. & Macklin, M.G. (1992)Alluvial sediment sequences in a glaciated catchment:the Voidomatis Basin, Northwestern Greece. EarthSurface Processes and Landforms 16, 207–26.

Woolfe, K.J. & Larcombe, P.I. (1998) Terrigenous sediment accumulation as a regional control on thedistribution of reef carbonates. In: Reef and CarbonatePlatforms in the Pacific and Indian Oceans (Eds G.F. Camoin & P.J. Davies), pp. 295–310. BlackwellScience, Oxford.

Woolfe, K.J., Larcombe, P. & Stewart, L.K. (2000) Shelfsediments adjacent to the Herbert River delta, GreatBarrier Reef, Australia. Australian Journal of EarthSciences 47, 301–8.

World Commission on Dams (2000) Dams and Develop-ment: a New Framework for Decision-Making. TheReport of the World Commission on Dams. AnOverview, 16 November 2000. http://www.dams.org(accessed 25 June 2004.)

ESD01 9/5/06 3:09 PM Page 426

REFERENCES 427

Wright, L.D. & Coleman J.M. (1971) The discharge/wave power climate and the morphology or delta coasts.Association of American Geographers Proceedings3, 186–9.

Wright, L.D. & Coleman, J.M. (1972) River delta mor-phology: wave climate and the role of the subaqueousprofile. Science 176, 282–4.

Wright, L.D. & Short, A.D. (1984) Morphodynamicvariability of surf zones and beaches: a synthesis.Marine Geology 56, 93–118.

Wright, L.D. (1982) Deltas. In: The Encyclopedia ofBeaches and Coastal Environments (Ed. M. Schwartz),pp. 358–68. Hutchinson-Ross, Stroudsberg, PA.

Wright, L.D. (1995) Morphodynamics of Inner Con-tinental Shelves. CRC Press, London, 241 pp.

Wright, V.P. & Burchette, T.P. (1996) Shallow-watercarbonate environments. In: Sedimentary Environ-ments: Processes, Facies and Stratigraphy, 3rd edn(Ed. H.G. Reading), pp. 325–94. Blackwell Science,Oxford.

Wu, J.S., Holman, R.E. & Dorney, J.R. (1996) System-atic evaluation of pollution removal by urban wetdetention ponds. Journal of Environmental Engineering122, 983–8.

Xie, S., Dearing, J., Bloemendal, J., et al. (1999)Association between the organic matter content andmagnetic properties in street dust, Liverpool, UK. TheScience of the Total Environment 241, 205–14.

Xue, C. (2001) Coastal erosion and management ofMajuro Atoll, Marshall Islands. Journal of CoastalResearch 17, 909–18.

Yair, A. (1994) The ambiguous impact of climate change at a desert fringe: northern Negev, Israel. In: Environmental Change in Drylands: Biogeo-graphical and Geomorphological Perspectives (EdsA.C. Millington & K. Pye), pp. 199–227. Wiley,Chichester.

Yair, A. & Gerson, R. (1974) Mode and rate of escarpment retreat in an extremely arid environ-ment (Sharm el Sheikh, southern Sinai Peninsula).Zeitschrift für Geomorphologie, Supplementband 21, 202–15.

Yair, A. & Lavee, H. (1976) Runoff generative processand runoff yield from arid talus mantled slopes. EarthSurface Processes 1, 235–47.

Yan, P., Shi, P., Gao, S., et al. (2002) 137Cs dating oflacustrine sediments and human impacts on DalianLake, Qinghai province, China. Catena 47, 91–9.

Yap, H.T. (2000) The case for restoration of tropicalcoastal ecosystems. Ocean and Coastal Management43, 841–51.

Yeats, P.A. & Bewers, J.M. (1982) Discharge of metalsfrom the St. Lawrence River. Canadian Journal ofEarth Sciences 19, 982–92.

Yechieli, Y. & Wood, W.W. (2002) Hydrogeologic pro-cesses in saline systems: playas, sabkhas, and salinelakes. Earth-Science Reviews 58, 343–65.

Young, R.A. (1985) Geomorphic evolution of theColorado Plateau margin in west-central Arizona: atectonic model to distinguish between the causes ofrapid, symmetrical scarp retreat and scarp dissection.In: Tectonic Geomorphology (Eds M. Morisawa &J.T. Hack), pp. 261–78. Allen & Unwin, Boston.

Yunker, M.B., Macdonald, R.W., Vingarzan, R., et al.(2002) PAHs in the Fraser River basin: a criticalappraisal of PAH ratios as indicators of PAH sourceand composition. Organic Geochemistry 33, 489–515.

Zhu, B.Q., Chen, Y.W. & Peng, J.H. (2001) Lead isotopegeochemistry of the urban environment in the PearlRiver Delta. Applied Geochemistry 16, 409–17.

Zimbelman, J.R., Williams, S.H. & Tchakerian, V.P.(1995) Sand transport paths in the Mojave Desert,southwestern United States. In: Desert Aeolian Processes(Ed. V.P. Tchakarian), pp. 101–29. Chapman andHall, London.

Zimmermann, M. (1990) Debris flows 1987 in Switzer-land: geomorphological and meteorological aspects.In: Hydrology in Mountainous Regions. II – ArtificialReservoirs; Water and Slopes (Eds R.O. Sinniger &M. Monbaron), pp. 387–93. IAHS Publication 194,Proceedings of two Symposia held at Lausanne, August.International Association of Hydrological SciencesPress, Wallingford.

Zimmermann, M. & Haeberli, W. (1992) Climate changeand debris flow activity in high-mountain areas – acase study in the Swiss Alps. Catena Supplement 22,59–72.

Zolitschka, B. (1998) A 14,000 year sediment yieldrecord from western Germany based on annually laminated lake sediments. Geomorphology 22, 1–17.

ESD01 9/5/06 3:09 PM Page 427

abyssal plain 354accommodation space 86, 107, 169, 171, 178,

264–5, 269, 300, 352acid mine drainage 5, 81, 98activation thresholds 84active intervention, coastal environments 283aeolian bedforms (arid environments)

barchan dunes 155, 162, 163, 164, 165, 166barchanoid ridges 162, 163, 165, 166blowout 163, 164, 165complex dunes 163compound dunes 163encroachment of sand dunes 184free dunes 163–5, 163impeded dunes 161, 163, 164, 165linear dunes 162, 163, 164, 165longitudinal dunes see linear dunesmobile dunes see free dunesnebkha 164parabolic dune 164, 165seif dunes 165simple dunes 162, 163, 163, 163, 164star dunes 162, 163, 164, 164, 165, 166transverse dunes 162, 165vegetated dunes 165

aeolian dust transport 155, 167aeolian processes (arid environments) see arid

environmentsaeolian processes (coastal environments) see dune

systems (coastal)aeolianite 151, 299aggregate dredging 290, 293, 335, 373–5, 380–1,

381, 385–6Agulhas Current 366, 366agricultural runoff 6, 81, 228, 243, 337Aire–Calder river, UK 91, 102, 197–9, 198,

209Akosombo Dam, Ghana 248

Index

Note: page numbers in bold refer to tables, those in italics refer to figures and boxes.

algae 111, 117, 128, 130, 135, 136, 139, 140, 229,304, 315, 342

algal blooms 252algal mats 312, 315, 323, 324, 327alluvial fans 45, 73, 77, 88, 89, 144, 155, 157,

159–61 160, 169, 171, 178, 186, 188alluvial plains 36, 49, 309, 310Amazon River, South America 76, 87, 303, 326,

326, 355, 362, 363, 364Amazon–Guiana Coast 30Amazon Shelf 326–7, 351, 354, 362Amoco Cadiz 238anaerobic diagenesis 14, 122, 211, 211–13, 337,

339, 340anhydrite 4, 150Andorra 32Aquatic Environmental Monitoring Reports 380Arabian Gulf 4, 356aragonite saturation state 304, 305, 306, 309, 316,

346Argentina River Delta, Italy 229arid environments

aeolian bedforms see aeolian bedforms (aridenvironments)

alluvial fans see alluvial fansanthropogenic impacts 182–4aridity see aridityCamanchaca see arid environments, coastal fogchannels see ephemeral streamsclimate impacts 187–9coastal fog 147definitions 144–6dew 148erosion 150–1, 179–81, 180, 185geographical distribution 145, 145hazards 184–7, 185, 186, 187landforms 156playas 157, 157

ESD02 9/5/06 3:10 PM Page 428

INDEX 429

pluvial lake 174–8, 175, 176, 177runoff 179–81, 180, 182, 182sediment

accumulation see aeolian bedforms (arid environments)

sources 150–1transport by water 151–4, 154, 179–181, 180transport by wind 154–6

soils 167, 167, 168tectonics 148, 169, 171–4, 172ventifacts 156weathering 149–50yardangs 156

aridityanthropogenic causes 148climatic causes 146–8continentality 147rain shadow 147tectonic causes 148

Ash Wednesday Storm 289Aswan Dam, Egypt 152, 249–51Atacama Desert, Chile 147–8, 168, 169, 189atmospheric

CO2 see climate and climate change, atmosphericCO2

deposition 6, 129, 193, 200temperatures 24, 307, 346

atoll see coral reefs, atollAvicennia 307, 311, 319, 324

Back-barrier 268–9, 278–9, 284, 299Bacterial-mediated reactions 14–5, 116, 122,

211–13, 211, 212, 215–18, 339, 346Bahama Bank 316, 317Bahamas 356Bahiá las Minas, Panama 341–2, 341Baja California 18–9, 19Barbados 342, 343baroclinic circulation 321, 322, 322barrier islands

definition of 263–4, 264, 268–9, 269morphology 18, 264, 266, 268–9, 268, 269, 270,

309, 310prograding 278regressive 25, 278storm impacts on 279–82, 280wave-dominated 277wave–tide dominated 277

barsdelta-mouth 268linguoid 78longitudinal 78, 159, 160

nearshore 272, 276, 277point 78transverse 78, 159, 276whaleback 63

Bay of Fundy, Canada 234beach

back-beach 274, 279, 283beachridge 267, 267, 270beach state models 276, 276, 277berm 272boulder 267, 275carbonate 268cusps 276, 277definition of 263dissipative 274, 276, 276, 277drift-aligned 275, 275erosion 18, 262, 265, 268, 279, 290–1, 360gradients 274–5, 275, 276gravel 264, 272, 274, 276, 279human impacts on 283–93lowering 284, 287management 293–8morphodynamics 273morphology 268–270, 269, 270, 274–83progradation 278reflective 276, 276recharge 30sediment

supply 266–8, 290–3transport 13, 271–4, 272, 273

storm reworking 279–83beachface 272, 274, 276, 277, 299beachrock 299bedforms see aeolian bedforms

see rivers, bedformsbedload 9, 16, 39, 40, 44, 152, 156, 158–9, 268,

273, 319, 321–2, 331, 355, 366, 372Bengal Basin 23, 87benthic flux 15, 213benzene 196berm, see beach, bermbioaccumulation 238bioavailability 29, 112Biodiversity Action Plan 378bioerosion 311, 312, 313, 314, 333, 336, 338, 347,

348, 356bio-irrigation 272biological oxygen demand 236, 259biological time-scales 1biomagnification see bioaccumulationbioremediation 346biotopes 375

ESD02 9/5/06 3:10 PM Page 429

430 INDEX

bioturbation 112, 116, 119, 120–2, 122, 124, 232,254, 312, 315–16, 318, 354, 357, 359, 361,362, 367, 372, 379

Birmingham, UK 204, 205, 214, 221Blackwater Estuary, UK 224blowouts see dune systems (coastal), blowoutsbrackish 15, 132, 213, 224, 226, 227, 231, 248,

248, 254, 260, 261Bradford Beck, UK 197–9, 199, 209Brazil 204Brazilian Shelf 356breaker zone 271, 272breakwater 285–7, 285, 288, 290bromine 195Bruun Rule 245, 297, 299, 300bryozoans 311, 312, 313, 316, 356Burdekin River, Australia 16, 17, 227, 355Buzzard’s Bay, USA 238, 252

cadmium 27, 81, 100, 119, 129, 383calcareous algae 268, 302, 305, 311, 312, 313, 315,

316, 317, 365calcification see coral calcificationCalifornian Coast 18, 291–3, 292Californian Shelf 354, 368Callianassa 315Canadian Environmental Protection Act 380,

383canal 29, 191, 191, 192, 193, 196–200, 201, 202,

210–11, 213, 214, 214, 215, 215–18, 215,221

Cane Bay, St Croix 330Cape Cod, USA 373Cape May, USA 287carbonate mud 314, 315, 320carbonate mudbanks 309, 315carbonate sediments see sediment, carbonateCaribbean 13, 341, 348, 358Carson River Basin, USA 104Catchment Management Plan 258, 258chemical grain dissolution 315, 318, 320chemical oxygen demand 208, 220chenier plain 333chenier ridge 325, 328, 333, 359–62, 360,

361Chesil Beach, UK 270, 293chloride 195, 205chlorophyll-a 111, 111, 130chromium 27, 81, 195, 219Clean Air Act 21cliffs 2, 2, 18, 25, 242, 266, 267, 267, 269, 284,

285, 286, 290, 325, 328

climate and climate changeair temperature 24, 307, 346arid environments 148, 187–9atmospheric CO2 20, 24, 189, 346, 347, 350desertification see desertificationENSO cycles 25, 85, 298models 23–5, 30, 189sea-surface temperatures 24, 333, 346–7, 347UV-radiation 20, 346

Clutha River, New Zealand 363, 372coarse braided rivers 40coarse debris system 37, 38, 45coastal

cells 274defence 25, 233, 243, 245, 248, 252, 255, 258,

260, 294dunes see dune systems (coastal)embayments 309, 368erosion 18, 25, 96, 248, 260, 266, 285–7, 286,

290, 291–3, 292, 298, 328, 331, 333, 335,343–5

land claim 225, 239, 241–2, 248, 252, 255–7,256, 260

morphology 264, 274–83, 300progradation 267squeeze 18, 25, 239, 245, 246, 248, 260–1storm-wave dominated coasts 263, 264swell-wave dominated coasts 264systems 25, 263, 266, 268, 273, 283, 295, 298,

343zone 1, 266, 268, 271, 293, 344, 354zone management 253, 255–9, 258, 262, 283,

295–8coastal barrier 2, 16Colorado River, USA 18–19, 19, 175Columbia River, USA 54, 55, 63, 368, 368combined sewer overflow (CSO) 197, 200, 209, 217combustion, of fossil fuels 5, 21, 81, 129, 193,

194–5, 196, 203Congo River, Congo 87contaminant

definition 5diffuse source 5, 81, 91, 99dissolved 5, 13, 94, 98, 99, 114, 195, 339, 342metals 5, 13, 15, 16, 28, 80, 81, 92, 94, 94, 98,

99, 102, 118, 130, 131–2, 142, 194, 194–5,203, 204, 205, 206, 208, 209, 212, 213,215–18, 236, 237, 237, 253–4, 254,339–40, 340, 371, 382, 383

organic 12, 15, 26–8, 81, 98, 132, 238, 196, 196,200, 208, 340, 346, 372

particulate 99, 236, 236, 314

ESD02 9/5/06 3:10 PM Page 430

INDEX 431

portioning 13, 80, 98–9point source 5, 81, 99, 112, 113, 208, 209, 255radionuclide 5, 28, 80, 81, 81, 98–9, 102–3, 103,

129–30, 130, 141, 142sources 5, 81, 193–200, 196, 201transport 99

continental shelf environmentsacross-shelf transport 357along-shelf transport 16, 331–2, 357, 357, 359,

363, 367anthropogenic inputs to 371–2climate change impacts on 379, 386–7current-influenced 364definition of 351distribution 351, 353, 357human impacts on 372–7management 378–84natural disturbances 357–70ocean current changes 387river-influenced 362–4sea-level change 387sediment

contamination 380extraction 373–5, 374, 375mixing 354, 357, 359, 361, 362, 367, 372, 379production (biogenic) 355–7remediation 379sources 267, 355–7transport 351, 353, 357–9, 360, 362, 362–9,

372–4, 375, 380, 383settings 354–5storm-influenced 358–62tide-influenced 357–8trawling impacts 353, 372–3, 373, 378–9, 385wave-influenced 362

continental slope 14, 351, 354, 372Convention for the Protection of the Marine

Environment of the North East Atlantic 371,378

convergent margin 354Cooper Creek, Australia 152copper 27, 81, 108, 118, 129, 195, 204, 210,

219–20, 219, 253coral atoll see coral reefs, atollcoral calcification 306, 346, 347Coral Creek, Australia 323, 324coral reefs

atoll 304, 309, 344, 348bank-barrier 309barrier 309, 349carbonate production 304, 316, 333, 335, 336,

342–3, 345–7, 350

classification 304, 309climate change impacts 346–50, 347corals 304–306, 306, 346, 355–6definition 303distribution 302–6, 303, 305, 306fringing 309, 316, 317geomorphology 305, 308, 309growth rates 348, 349human impacts on 333–343, 338, 340, 343management 343–6natural disturbance 325–33, 331, 332, 334nutrient impacts 336–9, 338oil impacts 339–40patch reefs 309phase shift concept 343reef flat 315, 331, 333, 348sea-level change 347–50, 347, 349sediment

budgets 313–14, 344, 348contamination 339–40, 340facies 316, 317

impacts of 335–6mixing 312, 315–16production 311–14, 312, 313remediation 345–6transport 312, 314–15, 325, 329–33, 330,

331, 332trapping 312, 315

storm impacts 329–33, 330, 331zonation 305

coralline algae see calcareous algaecreeks see tidal, creeks and channelscrevasse splays 78137Cs 81, 81, 82, 102–3, 103, 137, 137, 142, 204,

354, 369cuestas 36, 150currents

density 10, 11, 12littoral 288longshore 16, 272, 272, 326ocean 23, 147, 366, 387rip 272, 272secondary 272shore-parallel see currents, longshoretidal 269, 277, 279, 319, 322, 357, 363, 364,

365, 367traction 10, 13turbidity 10, 12, 14, 113, 114, 119, 120, 120,

128upwelling 147, 303, 305, 308, 309, 333, 366

wind-driven 329, 332, 352, 358, 364, 367, 377cuspate foreland 269, 269, 270

ESD02 9/5/06 3:10 PM Page 431

432 INDEX

cyclone 16, 25, 57–9, 58, 59, 264, 279, 280–2, 325,328, 328, 329–33, 331, 337, 343, 344, 347,352, 353, 357, 358, 359, 359–62, 360

Cyclone Joy 332Cyclone Winifred 359

dammed lake 13, 49, 61Darwin Harbour, Australia 310, 319DDT 132, 132, 238debris avalanches 46, 46, 48, 54, 55, 56, 57, 63debris flows 14, 14, 38, 41, 45, 46–7, 46, 47, 48, 48,

50, 51, 52–4, 57, 59, 60, 60, 63, 64, 66, 67,69–70, 73, 89, 107, 158, 159, 160, 161,174–8

Dean’s Parameter 277Dee Estuary, UK 225deforestation 32, 57, 64–6, 65, 79, 81, 81, 90–2,

95, 105, 336, 354, 355Delft, Netherlands 200, 211delta (coastal)

birdsfoot 227–8classification of 225–8, 227climate change, impacts on 260–2channels 223, 232, 248, 260cuspate 227–8, 227definition of 223–4distributary channels 223, 228, 249, 260ebb-tidal 268, 269, 277–8, 279, 285, 287erosion 239–43, 240, 241, 243evolution 243–5, 244flood-tidal 268, 285fluvial-dominated 227–8, 227front 228, 229, 230, 235, 239, 239, 243, 245, 249human impacts on 247–52, 250, 250lobate 227–8, 227lobes 244prodelta 230, 231river-dominated 227–8, 227sea-level change, impacts of 260–2sediment

accumulation 229–235, 230anthropogenic inputs 235–8, 236, 237contaminant reworking 252–5supply to 228–9, 228transport 229–35, 233

tidal 264, 268, 269, 269, 278–9tide-dominated 227–8, 227

dentrification 337, 339desert pavements 167, 168desertification 79, 148, 179, 184, 187, 188deserts 144–6developed coasts 266

Devensian 86Diadema 311, 313, 337, 342diagenesis 2, 14–15, 30, 122–3, 201, 202, 208, 211,

211, 212, 213, 215–18, 237, 274, 299diatoms 5, 229, 314, 356digital terrain model 169dimictic lake 112dioxins 81, 132, 132, 133–4, 133, 196dissolved oxygen 367distribution coefficient, in lakes 127–8, 129docks see canaldrainage

densities 24, 58networks 94, 98, 171, 207, 220

dredged material disposal 211, 354, 371, 380–3,381, 383, 384

dredging see sediment, dredgingsee aggregate dredging

drift-aligned see beach, drift-aligneddrowned valley, see estuaries, evolutiondrylands 144–7, 157, 185, 188, 189Duchesne River basin, Utah 97, 97dune systems (coastal)

agricultural impacts on 290blowouts 269, 270, 290definition of 269development of 264evolution 270, 275–9foredunes 269, 270, 272, 274, 278, 282–3, 290human impacts on 283–90, 290–3management of 293–295morphologies 269, 270parabolic 269, 270recession 278, 282, 300removal 266, 283, 291sea-level rise, impacts on 299, 300sediment

aeolian transport 9, 9, 13, 273sources 264–5supply 274

transgressive 267, 269, 270transverse 270vegetation 265, 269, 270, 274, 290, 298vegetation destabilization 283–4, 290

dunes (arid-zone and desert) see aeolian bedformsduricrust 169, 170dynamic equilibrium 83

East Australian Current 365–7, 365East China Sea 362ebb delta see delta, tidalecosystem indices 137, 139, 141

ESD02 9/5/06 3:10 PM Page 432

INDEX 433

ecosystem-based management 384effect–load–sensitivity analysis 139, 141–3, 143El Niño Southern Oscillation (ENSO) 25, 85, 189,

298, 325, 333, 344, 347endorheic lake 24Environmental Consequence Analysis 137–41Environmental Impact Assessment 381Environmental Impact Statement 377Environmental Quality Objective 259Environmental Quality Standard 253, 259environmental sedimentology, definition 1ephemeral streams

bedforms 159flash flood 153flow characteristics 152–4, 152, 153, 154hydrograph 153sediment storage 157–9suspended sediment 152, 157–9, 158transmission losses 154, 155

epicontinental shelves 354epilimnion see lakes, epilimnionerosion

channel bank 6, 44, 81, 83, 84, 90, 92, 94, 107,197, 200, 318

coastal see coastal, erosionexposed lake sediment 144, 151hillslope 87, 171–4, 186

estuariesclassification of 225–7, 226definitions of 223–4evolution 223–4human impacts on 248–52, 261–2hydrodynamics 268management 255–9, 258mixed 227, 228partially mixed 226–7, 226salt-wedge 226–7, 226, 231, 235, 239sea-level rise, impacts of 245–7, 260–1sediment

accumulation 229–35anthropogenic inputs 235–8deposition 224–5sources 228–9mixing 225–7, 226transport 229–35, 233

stratified 227, 227Estuary Management Plan 257, 258, 261Euphrates, Turkey/Syria/Iraq 76, 152eustatic 88, 325eutrophication 110–12, 111, 111, 114, 123, 124,

126, 130, 130, 132, 134, 136, 139–41, 140,142, 209, 337, 342, 343, 349

evaporation 4, 24, 129, 132, 147, 157, 170, 185,249, 304, 311, 341

evaporites 113, 170evapotranspiration 146, 147, 178, 321exorheic 24

Fal Delta, UK 229flocs see flocculationflocculation 12, 28, 131, 231, 232, 235, 318, 321,

322flood delta see delta, tidalflooding

arid environments see ephemeral stream, flashflood

outburst 51, 51rivers see rivers, floodingsewer 208

floodplaincontaminant deposition and storage 92, 93, 94,

99, 102–4, 103contaminant remobilization 104sediment deposition 78, 78, 83, 83, 84sediment profiles 104

Florida Keys, USA 337, 338, 338Florida Reef Tract, USA 337–8flow

debris 14, 14, 38, 45–7, 46, 47, 48, 50, 51, 52–4,57, 59–60, 60, 62, 63, 66–7, 69, 73, 89, 107,158, 159, 160, 161, 175, 178

grain 14, 14granular 46, 46hyperconcentrated 46, 47, 48, 49, 57laminar 7, 8, 47, 71, 127, 131microturbulent 322turbulent 7, 8, 11, 47, 127, 322

fluiddensity 7flow 7–8, 47viscosity 7–8, 9, 10

fluvial see riversfluvioglacial sediment 13, 49–51Fly River, Papua New Guinea 108, 310, 319Follets Island, USA 280–2, 281, 282foraminifera 268, 302, 305, 311, 312, 313, 315,

316, 317, 318, 320, 337–8, 356foreland basin 354foreshore 273forest fires 79, 81, 84, 88, 89, 107French Guiana 325, 326–7, 326frost 24, 45, 306, 307Froude number 8, 8, 10–11Futurecoast 297, 301

ESD02 9/5/06 3:10 PM Page 433

434 INDEX

Galveston Island, USA 280–3, 281, 282Ganges River, India 23, 64, 74, 76, 86Ganges–Brahmaputra Delta, India 227, 325geoindicators, of environmental change 296, 338George River Delta, Canada 229German Bight 321Gila River, USA 153, 155glacial

ice 13, 48, 177moraine 13, 38, 44, 49, 66, 70systems 48–51till 13, 45, 66, 67, 285, 286

Glenn Canyon Dam, USA 18global climate models 23, 30–1, 189Gobi Desert, Inner Mongolia 145, 155grain shape 6, 6–7, 79, 156grain size 6, 6–7, 7, 9–10, 9, 12–13, 40, 48, 63, 82,

99, 114, 117, 119, 119, 120, 126, 151, 156,205, 221, 224, 225, 230, 231, 274, 275, 276,277, 282, 288, 314, 315, 321, 336, 358, 362,368–9, 373, 374, 376

gravel beaches see beach, gravelGreat Bahama Bank 316, 317Great Barrier Reef, Australia 308, 336, 339, 345Great Barrier Reef Marine Park Authority 345,

354Great Barrier Reef shelf 16, 316, 330, 331–2, 332,

351, 357, 359–62, 360, 364, 365, 368, 370,383, 385

greenhouse gases 24Grijalva–Usumacinta Delta, Mexico 310, 323,

324groynes 30, 267, 283–4, 285–7, 285, 286, 288,

294, 295, 296Guiana Current 326, 326, 363, 364Gulf of California 18–19, 19Gulf of Mexico 177, 244, 330, 338, 354, 358Gulf of St Vincent 323, 324gully pot 205–8, 207gypcrete 168, 169, 170, 183gypsum 4, 150, 161, 167, 170, 173, 187

Habitats Directive 382Halimeda 312, 313, 313, 314, 315, 316, 317, 338Hallsands, UK 293halophytic vegetation 224, 247Hamburg (port), Germany 26–8, 26, 27Hanalei Bay, Hawaii 14Hawaii 205, 220, 221, 314Hayasui Strait 357heavy metals see contaminant, metalsheavy minerals 3, 10, 82, 293, 362, 368

Herbert River, Australia 369, 370herbicides see contaminants, organicHimalaya 23, 64, 65, 65, 74, 82, 87, 96, 148Himalayan environmental degradation theory see

HimalayaHjulström curve 7, 7, 9, 119Holocene 23–4, 25, 45, 67, 70, 85, 86, 87, 148,

174–8, 177, 188, 298, 300, 314, 325, 331–2,348, 349, 350, 352, 359, 361, 362, 372, 375,387

Holocene Climatic Optimum 298Holothurians 312, 315homopycnal 228, 234Hoover Dam, USA 18, 95Huascaràn, Peru 46, 52–4, 52, 53Humber, UK 229, 245, 270humic state index see lakes, humic statushurricane see cycloneHurricane Andrew 331Hurricane Carla 358Hurricane Hugo 330Hurricane Isabell 280hydrocarbon production 371hydrocarbons see contaminant, organichydroelectric power 32, 95, 249, 372hydrograph (flood) 17, 153, 197hydro-isostacy 355hydrological cycle 20, 24hyperconcentrated flow see flow,

hyperconcentratedhyperpycnal 228, 235hypersynchronous 235hypertidal 226hypertrophic lake see lake, hypertrophichypolimnion see lakes, hypolimnionhypopycnal 228, 235hyposynchronous 235

ice caps 13, 24, 49Indian Ocean tsunami 333, 334Indonesian Archipelago 355Inhaca Island, Mozambique 307, 308, 309, 320,

329Irish National Sea Bed Survey 384Irish Sea 285, 285, 384iron oxides 14, 193, 195, 203, 203, 212iron reduction 211, 212, 212isostatic uplift 299isotherm 303, 304, 306

Jamaica 308, 309, 316, 317, 339, 340, 342Jökullhaup 13, 47, 48, 49

ESD02 9/5/06 3:10 PM Page 434

INDEX 435

Kailua Bay, Hawaii 314Karakorum, Pakistan 34, 35, 64, 66, 66Kärkevagge, Sweden 37, 38, 43, 45, 45Key Largo, Florida 338Kosi Lagoon, South Africa 279

La Niña 25, 85, 189Lacepede Shelf 351, 356, 362lagoon

coastal 260, 264, 264, 267, 268, 269, 269, 278,278, 279, 305, 309, 310, 323

reef 304, 305, 312, 314, 315, 316, 317, 331, 348

lahar 62, 63, 89, 90Lake Batorino, Belarus 134, 135–6, 135Lake Burdur, Turkey 151Lake Ekoln, Sweden 137, 138lake pollution see lakes, sediment, pollutionlakes

bioturbation 120–2, 120, 122bottom dynamic conditions 116–20, 118, 118,

120classification 109, 110–12, 110, 111, 111climate change, impacts on 132–4, 135–6, 135colour 111–12, 111definition 109deltaic areas 113diagenesis 115distribution coefficient 127–8earthquake records 128epilimnion 112eutrophic 110–11, 111eutrophication 139–41, 140fluctuations in lake level 128glacial 110, 110humic status 111–12, 111hypertrophic 110–11, 111hypolimnion 112management 137–43, 142, 143mass balance modelling 113, 125–8, 125, 125,

127, 128, 137, 142, 143, 143mass movement 128mesotrophic 110–11, 111minerogenic matter 114, 114monomictic 112polymictic 112oligotrophic 110–11, 111remediation 137–43, 142, 143sediment

accumulation 112, 116–19, 118, 118allochthonous 113, 114, 115autochthonous 113, 114, 115

chemical characteristics 123–5, 124classification 113–14, 114, 115compaction 116dating 134–7, 137erosion 116hydrogenous 113records of sedimentation 134–7, 137pollution 119, 128–32, 130, 131, 132, 133–4,

133, 138resuspension 112, 116–20, 118sources 113–14transport 116

slope processes 113suspended particulate matter 114–16, 115thermocline 118–19trophic status 110–11, 111toxicity 130–2, 130, 132turbidites 113types see lakes, classification

laminar flow see flow, laminarland claim see coastal, land claimlandslide 25, 46, 47, 47, 48, 51–2, 52–4, 57–9, 58,

59, 62, 63, 64–5, 66, 67–8, 70, 85, 90, 91,107, 108, 110, 128, 171–3, 172, 200, 267

Last Glacial Maximum 70, 177, 355lead (Pb) 15, 21–3, 22, 81, 92, 98, 99, 124, 130,

131, 184, 191, 194–5, 194, 196, 196, 203,204, 205, 206, 209, 210, 214, 215, 220, 221,221, 237, 253, 383

levee 45, 48, 54, 56, 78, 78, 95, 161, 239, 310, 324Little Ice Age 59, 60, 60, 85, 86, 298littoral cell 292littoral sediment budget 292, 348LOIS programme 197, 209longshore current see currents, longshorelongshore drift 250, 264, 266, 267, 275, 275, 279,

281, 285, 287, 295Los Angeles, USA 151, 291–3, 292Los Angeles Aquaduct, USA 151

Mahakan Delta, Indonesia 229managed realignment 252, 253–4, 255, 262Manchester, UK 193, 194–5, 194, 203, 204, 205,

206, 210, 214, 215–18, 215, 217manganese 13, 14, 15, 98, 114, 115, 117, 118, 124,

124, 130, 131, 131, 195, 196, 196, 201, 204,213, 216, 217, 218, 340

manganese reduction 211, 212, 212, 213, 217, 218mangroves

classification 309, 310climate change, impacts on 346–50, 347crabs 319–20

ESD02 9/5/06 3:10 PM Page 435

436 INDEX

mangroves (continued )creeks 11, 318, 321–2, 322definition 306distribution of 303, 306–9, 307erosion 325–8, 328, 329flats 321–2geomorphology 307, 309–11, 310human impacts on 333–40, 341leaf litter 319, 323management 343–6nutrient impacts 336–9oil impacts 339–42, 341pneumatophores 323, 326, 340root networks 320, 323, 336, 339sea-level change 347–50sediment

accumulation 318, 323contamination 339–40facies 323, 324impacts of 336mixing 318, 323production 318, 319–20, 320remediation 345–6supply to 319transport 318, 320–2, 322trapping 318, 322–3

shoreline dynamics 325–9, 326, 327, 328storm impacts on 329–33vegetation 307, 311, 319zonation 307, 311

marine aggregates see aggregate dredgingmarine mineral extraction see placer depositsmarsh see salt marshmass balance modelling see lakes, mass balance

modellingmass movement 39, 45, 46, 46, 47, 47, 52, 60,

62–4, 66, 68, 69, 73, 117, 128, 173Mediterranean 75, 179–81, 183, 189, 265, 351,

373megaclasts 48, 333Mekong Delta, Vietnam 227, 325mercury 27, 81, 94, 129, 130, 137, 138, 143, 383mesa 150mesotrophic lakes see lakes, mesotrophicmetals see contaminant, metalsmethane 212, 213, 217, 346methanogenesis 15, 211, 212, 213microfossils 5, 282, 356mid-Atlantic Bight 351, 373, 373Middle Creek, Australia 322mine waste 5, 80, 93, 99, 100–1, 291

mineral magnetics 3, 82, 193mining

acid mine drainage 5, 81, 98channel and floodplain aggradation 92metal contamination see contaminant, metalsriver, impacts on see rivers, miningwaste 92–4, 93

Mississippi, River, USA 76, 86, 95, 226Mississippi Delta, USA 227, 228, 230, 235, 238,

243, 244, 249, 257monsoon 17, 23, 64, 152, 177, 177, 359moraine see glacial, moraineMorecambe Bay, UK 242, 243, 244Mont Blanc, France/Italy 67, 69Mount Rainier, USA 47, 49, 50, 51, 51Mount St Helens, USA 46, 47, 54–6, 55, 62, 63,

63, 64, 69mountain environments

climate change, impacts on 69–70climate-induced events 57–60, 58, 59, 60debris flow 38, 41, 45, 46–7, 46, 47, 48, 48, 50,

51, 52–4, 57, 59–60, 60, 63, 64, 66–7,69–70, 73

definition and classification 33–7, 34, 35, 36deforestation 64–6, 65glacier systems 48–51, 50, 51global distribution 35–7, 35, 36hazards and hazard monitoring 67–9hillslope processes 46–7, 47, 57–62, 58, 59, 61human impacts 64–7, 65, 66landslides 38, 41, 45, 46, 47, 47, 48, 48, 50,

51–2, 52–4, 57–9, 58, 59, 62, 63, 64–5, 66, 67–8, 70, 73

microscale modelling 70–1, 72rivers 40–3, 40road construction 66–7, 66sediment

budget 43–6, 43, 44, 45, 59flux 43, 43mountain sediment cascade 37–40, 38, 38,

39transfer 46–7, 46, 47yields 40–3, 41, 42

seismic-triggered events 51–7, 52, 53, 55volcanic-triggered events 54–6, 55, 62–4, 63

mudcarbonate mud 314, 315, 320fluid mud 232, 362mudbanks 326–7, 326, 327

mudflat 231, 232, 233, 236, 239, 245, 246, 246,247, 260, 261, 324, 326–7, 333

ESD02 9/5/06 3:10 PM Page 436

INDEX 437

Namib Desert, Namibia 145, 147, 148, 183Namibia 148, 293, 375, 376, 377Natal Coast, South Africa 308, 309Negev, Israel 148Nepal 39, 64, 65, 87New Caledonian Shelf 367Newfoundland Shelf 364Newtonian fluid flow see flow, turbulentNile Delta, Egypt 18, 229, 248, 249–51, 250, 250,

256, 260, 261, 262Nile River, Egypt 250nitrate 14, 15, 169, 183, 195, 212, 213, 217, 306,

306, 337non-Newtonian flow see flow, laminarNormanby River Estuary, Australia 310, 319North Atlantic Oscillation (NOA) 25, 298North Sea 26, 28, 354, 371, 372, 378Norway 32, 374nutrients 5, 14, 75, 80, 81, 91, 98, 99, 100, 108,

111–12, 117, 118, 124, 126, 127, 129, 132,142, 195, 200, 208, 212, 213, 221, 236, 247,248, 262, 268, 290, 304, 306, 306, 309, 333,336–9, 355, 367, 372, 378–9

ocean currents see currents, oceanoffshore windfarms 377, 386oil spills 238, 252, 341–2, 341oligotrophic lakes see lakes, oligotrophicooid 305, 314Orange River Delta, South Africa 375organochlorine pesticides see contaminants, organicOrinoco, Venezuela 3, 76, 227, 303, 326Otago Shelf 357, 362, 363, 372outwash fan 66, 67overwash fan 268, 272, 272, 274, 276, 279, 298Owens Lake, USA 151, 174Owens Valley, USA 151oxalate 204

Pahang Delta, Malaysia 229PAHs 27, 196, 196, 371, 383Panama 333, 341–2, 341, 347parabolic dunes see aeolian bedforms (arid

environments), parabolic dunesee dune systems (coastal), parabolic

paraglacial 48, 264, 299passive intervention 283passive margin 148, 354210Pb 82, 137, 184, 354, 363, 369PCBs 27, 28, 80, 132, 132, 196, 196, 209, 238, 255,

371, 383

pediment 150peloid 314Penicillus 312, 313, 315pericontinental shelves 354permafrost 59, 60, 60, 69, 70, 84permeability 168, 180, 182, 267, 320persistent organic pollutants see contaminant,

organicpesticides see contaminant, organicphosphorus 28, 81, 91, 110–11, 112, 117, 118,

122, 123, 124, 126–8, 128, 132, 135, 136,139, 140, 141, 200, 209, 219, 219, 221, 337

photic zone 116, 143, 305, 306placer deposits 375platinum 169, 204, 214, 374platinum group elements (PGE) 196, 196playa 144, 157, 175Pleistocene 67, 86, 87, 88, 171–4, 174–8, 175,

177, 249, 350, 375PM10 201PM2.5 201pockmarks 367point-bar see rivers, bedformspollutant see pollutionpollution, definition 5

see also contaminantPortuguese Island, Mozambique 328post-glacial transgression 325, 355, 375pro-glacial sediment sources 43, 44, 49pyrite 115, 213, 214, 327pyroclastic flows 57, 62, 63

Quaternary 23, 75, 88, 148, 149, 170, 174, 178,188, 332

radionuclides see contaminant, radionuclideRajang River Delta, Malaysia 16reefs see coral reefsremediation see sediment, remediationReynolds number 7–8Rhine–Meuse Delta, Germany 244Rhizophora 307, 311, 320, 324rhodoliths 365–6Ribble Estuary, UK 225Richmond River Estuary, Australia

319Rio Pilcomayo, Argentina 81, 92, 93Rio Tinto, Spain 80, 80, 98rip currents see currents, ripripples 8, 8

ESD02 9/5/06 3:10 PM Page 437

438 INDEX

river catchments 3, 3, 345River Elbe, Germany 26–8, 26River Ettrick, UK 80, 81River Odra, Poland 81River Ouse, UK 3, 3, 99, 107River Rhine, Germany 43, 43, 76, 200River Teviot, UK 80, 81River Thames, UK 86, 91, 226River Tweed, UK 80, 81River Tyne, UK 15, 81, 100rivers

afforestation 90–2agricultural impacts 90–2anastomosing 75–8, 77, 78bedforms 78, 78braided 75–8, 77, 78classification 75–8, 76, 76, 77, 78climate and climate change 75, 76, 84–7, 85,

86contaminants 80–1, 81, 98–104, 101, 103,

107dams 95–6, 96, 107definitions 75–6deforestation 90–2ephemeral see ephemeral streamsflooding 85–6, 86, 94, 96, 105, 107floodplain see floodplainforest fires 88, 89glaciostatic rebound 87–8global distribution 75, 76human impacts 89–98, 90, 93, 94, 96, 97management 104–6meandering 75–8, 77, 78mining 92–4, 93, 94, 108regulation 90, 94–7, 97, 107–8, 290restoration 104–6riffle 40, 78, 106sediment

characteristics 79–81fingerprinting techniques 80–2, 80, 82particle shape 79provenance 79–80, 80, 81–2, 82sources 79–82, 80, 81, 82supply 82–4, 84transport 82–4

straight channel 75–8, 77, 78tectonic uplift 87–8urban see urban environments, riversurbanization 98, 108volcanic eruptions 89

road dust see road-deposited sedimentroad salt 193, 196, 205

road-deposited sediment (RDS)accumulation 192, 202–5composition 202–5, 203, 204definition 191grain size 205, 221management 218–20, 219metals 194–5, 194, 196, 196, 203–5, 203, 204,

219metal speciation 194–5, 194, 204mineral magnetic analysis 193organic contaminants 196, 196sources 193–6, 193, 196spatial variability 205, 206street sweeping 218–20, 219temporal variability 203–4, 205, 214transport 200–2, 201, 202

rock fall 14, 14, 38, 39, 45, 45, 46, 47, 47, 51, 60,61, 62, 66, 67

rock slide 38, 46, 60, 61, 61, 62, 68Ross Creek, Australia 322Rosslare Harbour, Ireland 285–7, 285, 286Roundness 6, 79Roxburgh Dam, New Zealand 372runoff

agricultural 6, 81, 228, 243, 337urban 6, 20, 81, 98, 195, 197, 200, 202, 204,

218, 219, 228, 238, 243

sabkha 144, 156, 157, 170Sahara 13, 145, 145, 146, 148, 149, 156, 179, 188Saharan Shelf 364salcrete 169, 170, 274Salford Quays, UK 215–18, 215, 216, 217Salicornia 232, 242salients 269, 288saline incursion 321salinization 79salt marsh

contaminant cycling 252–5, 254creeks 234, 234erosion 239–43, 240, 243evolution 224–5land claim 225, 256management 255–7oil impacts 252–5sea-level rise, impacts of 244–7, 246, 260–1sediment

accumulation 232, 233–4sources 228–9trapping 247

vegetation 232, 238wave attenuation 233–4

ESD02 9/5/06 3:10 PM Page 438

INDEX 439

salt wedge see estuariessaltation 9, 9, 10, 13, 156, 209, 273San Francisco Bay, USA 15San Joaquin River, USA 81sand abrasion 185sand dunes see aeolian bedforms (arid environments)

see dune systems (coastal)sand mining 266Scolt Head Island, UK 234scree 14, 39, 67, 110sea ice 24, 263, 264sea surface temperature 20, 24, 303, 304, 305, 307,

309, 333, 346, 347, 347sea wall 2, 233, 241, 246, 247, 284, 287, 288, 294seagrass 304, 312, 313, 315, 316, 320sea-level change 2, 2, 24, 25, 75, 88, 225, 239, 244,

245, 246, 247, 253, 255, 257, 260, 260–1,262, 277, 283, 284, 289, 296, 297, 298–300,298, 325, 347–50, 349, 354–5, 387

sea-level fall see sea-level changesea-level highstand (Holocene) 25, 352sea-level rise see sea-level changesecondary circulation 318, 321secular time-scales 1sediment

arid zone see arid environments, sedimentbeach see beach, sedimentbedforms see aeolian bedforms (arid

environments)see barssee rivers, bedforms

biogenic sediments 1, 2, 4–5, 4, 20, 131, 229,337, 354, 355–7, 363, 379, 387

bioturbation see bioturbationbudgets 43–6, 44, 48, 81, 90, 92, 106, 202, 313,

314, 335, 384capping 379, 380, 382carbonate 4, 311–14, 312, 313, 315, 316, 337–8,

338, 339, 342, 356cells 278, 292, 380chemical treatment 29, 379contamination see contaminantcontinental shelf see continental shelf

environments, sedimentcoral reefs see coral reefs, sedimentdating see sediment datingdeltaic see delta (coastal), sedimentdredging 26–8, 29, 221dune see dune systems (coastal), sediment

see aeolian bedforms (arid environments)entrainment 7–9, 7, 9, 167, 315estuarine see estuaries, sediment

extraction 94–7, 248, 293, 344facies 316, 317, 323, 324, 353, 358, 362, 363,

364, 365–7, 366, 368–9, 387fingerprinting see sediment fingerprintingfluvial see rivers, sedimentflux 37–8, 41, 43–5, 43, 44, 75, 85, 105, 171,

197–9, 209, 275, 292, 296, 314, 335–6intertidal 16, 228–38, 232, 239–47, 255,

316–24, 318, 320, 324lake see lakes, sedimentmangrove see mangroves, sedimentmarsh see salt marsh, sedimentmountain and upland see mountain environments,

sedimentoffshore see continental shelf environments,

sedimentparting zones 357, 366pollution see contaminantremediation 29, 213–5, 215–8, 216, 221, 345–6,

381river see rivers, sedimentsalt marsh see salt marsh, sedimentsettling 10, 321size see grain sizesorting 6–7, 6, 40, 288, 316suspended 3–4, 3, 11–13, 17, 20, 35, 36, 39, 41,

41, 44, 55, 57–9, 59, 80, 82–3, 85, 87, 92,95, 99, 106, 152, 157–8, 169, 183, 187,197–9, 198, 199, 201, 202, 208, 209, 210,220, 247, 314, 319, 321, 322, 336, 340, 364,368, 373

terrigenous inputs 268, 314, 323, 331–2, 332,336, 363, 368

texture 6–7, 6transport 7–14, 7, 8, 9, 11, 12, 14, 16, 17, 18–19,

19, 20, 46–8, 47, 48, 82–4, 116–19, 118,118, 151–6, 200–2, 229–35, 233, 271–4,272, 273, 314–5, 320–2, 357–69

traps 95, 109, 137, 296, 369treatment 27, see also sediment, remediationurban see urban environments, sedimentyields 24, 33, 35, 40–3, 55, 63–4, 91, 92, 209,

354, 355sediment dating 121, 134–7sediment fingerprinting 3, 3, 18–19, 80, 80, 81–2,

82, 197Sediment Quality Guidelines 28, 29, 221, 382Sediment Quality Triad 29Severn Estuary, UK 225, 226, 236, 237, 237,

239–42, 240, 241, 253–4sewage 81, 81, 98, 196, 197, 198, 200–1, 204, 209,

210, 211, 213, 217, 236, 259, 339, 371

ESD02 9/5/06 3:10 PM Page 439

440 INDEX

sewer systems see urban, sewer systemsshear stress 7, 8, 9, 90, 158, 165, 315shoreface 264, 267, 272, 276, 277, 278, 280–2,

289, 299, 331, 353shoreline

erosion 18, 25, 249–51, 250, 285–7, 328, 331,335, 344

management 257–9, 293–8overstepping 298, 298progradation 2, 2, 267, 277–8, 278, 295, 325,

328protection 29–30, 283–90, 288, 289, 344rollover 298–9, 298transgression 300, 325

Shoreline Management Plan 258, 258, 295side-scan sonar 369slack water 11, 231slope failure 25, 57–62, 98, 267slope instability 33, 67–8, 69, 70soil 76, 79, 80, 81, 85, 88, 112, 124, 128, 129, 131,

139, 148, 167, 167, 169, 183, 193, 193, 196,262

soil creep 38soil erosion 6, 25, 64, 65, 65, 66, 84, 86, 90–2, 95,

108, 182, 183–5, 200, 229, 335South Africa 279, 293, 299, 303, 309, 375South Otago Shelf see Otago Shelfsouth-east African Shelf 366Southern Ocean 362Southland Current 362, 363Spartina 232, 238, 242, 247spit 268, 269, 269, 270, 278, 278, 285–7, 285, 294step–pool systems 40Stoke’s Law 231storm

bed 353, 359–62frequency 2, 20, 25, 239, 290, 346, 349surge 330, 331, 333, 353wave 16, 25, 261, 263, 264, 265, 329, 340, 347,

365street sweeping see road-deposited sediment, street

sweepingsubglacial 13, 49subtidal 228, 232, 236, 263, 295, 330, 369sulphate reduction 15, 211, 212, 213, 339, 372sulphide 15, 93, 130, 237, 339supratidal 4, 263, 264, 293surf zone 271–4, 272, 279suspended particulate matter see lakes, suspended

particulate mattersustainable drainage systems (SuDs) see urban

environments, sustainable drainage systems

swash zone 272swath-bathymetry mapping 369, 377, 378,

387Switzerland 38, 43, 44, 47, 59, 60, 61, 62, 67

talus deposit 60, 330, 331talus slope 66, 151Tarbella Reservoir, Indus River, Pakistan 187tectonic subsidence 18, 25, 96, 158, 171, 249,

251, 284, 344tectonic uplift 57, 66, 81, 84, 86, 87–8, 107,

148, 169, 171–4, 355tectonics 35, 40, 77, 147, 148, 151, 169–71,

171–4, 353, 355Tees Estuary, UK 255, 256Tejo River, Portugal 87terrain zones 39Texas–Louisiana Shelf 351, 357, 358, 359Thalassia 315Thaw Estuary, UK 225Tibesti Plateau, Chad 156tidal

barrages 248–52creeks and channels 232, 234, 321–2, 328currents 269, 277–8, 279, 284, 319, 321–2,

357, 363, 367, 386delta see delta (coastal), tidalflat 232, 322–3inlets 267, 268, 269, 277–279, 280, 284inundation 225, 245–7, 321, 344, 347, 350prism 225, 239, 277, 279, 284, 286, 287,

321, 335pumping 321–2

tideasymmetry 11, 11, 231–3, 235, 321–2cycle 11, 11diurnal 273, 321, 357ebb 10–11, 11, 224, 231–2, 232, 233, 235, 248,

318, 321–2, 323, 335flood 11, 11, 224, 225, 231–2, 232, 233, 235,

248, 260, 318, 321–2neap 252macrotidal 226, 235, 239, 253, 276, 319mesotidal 226, 319microtidal 225, 235, 319, 323range 225–6semi-diurnal 357spring 232, 246, 307, 321, 322

tombolo 110, 269, 269, 288Torres Strait 357toxicity, of pollutants and contaminants 29,

130–2, 238

ESD02 9/5/06 3:10 PM Page 440

INDEX 441

traction carpet transport 13travertine 5trawling see continental shelf environments, trawling

impactstributyl tin 133–134, 200trophic status index see lakes, trophic statustsunami events 16, 29, 275, 279–83, 325, 329–33,

334, 344, 353, 357, 367–8tsunami deposits 333, 368tufa 5, 174, 176, 188turbidite 12, 12, 14, 113, 114, 128turbidity 12, 208, 226, 247, 248, 302, 308, 336turbidity maximum 226, 322, 322typhoon see cyclone

Udden–Wentworth scheme 6, 6UN Convention on Biodiversity 378Upper Rhine, Germany 43, 43urban environments

anthropogenic impacts 213–18, 216canals see canalclimate change impacts on 222definition 190–1diagenesis 211–13, 211, 212, 214, 215–18,

217docks see canalgully pots 205–8, 207lakes 200, 210–11, 212rivers 196–200, 198, 199, 202, 208–9, 210sediment

environments 191, 192management 218–21, 221metal speciation 194–5, 194, 204metals 194–5, 194, 196, 196, 203–5, 203,

204, 209, 210, 212, 213, 214, 215–18,216, 217, 219

mineral magnetic analysis 193road-deposited sediment see road-deposited

sedimentsources 193–200, 193, 196, 198suspended sediment 197–9, 199, 209, 210transport 200–2, 201, 202

sewer systems 200–1, 201, 202, 205–8sustainable drainage systems (SuDs) 222urban sediment cascade 200–2, 201

URGENT programme 197, 209upwelling see currents, upwellingUV radiation 20, 346

ventifacts see arid environments, ventifactsvivianite 15, 115, 212, 213, 214Volta Delta 248

Wadden Sea 228Waipaoa, New Zealand 46, 57–9, 58, 59Wash, UK 225, 234washover 264, 290Water Framework Directive 30, 382water quality 106, 192, 195, 197, 208, 210, 211,

213, 215–18, 218, 219, 220, 221, 222, 252water resources 109, 151, 183, 260wave

attenuation 234base 11, 12, 112, 113, 118, 119, 120, 120, 121,

128, 271, 298, 305breaking (types) 276, 276dampening 233–4diffraction 271edge waves 271, 272energy 4, 227, 233, 234, 244, 264, 265, 271, 272,

272, 273–4, 275, 276, 277, 278, 279, 287–8,290, 386

height 12, 261, 264, 272, 276, 277, 329, 333incident 271, 272, 276, 276, 277infragravity 279, 357length 12, 271motion 12, 277overtopping 25, 261, 347reflection 272refraction 271, 272, 288, 293sea 12, 265storm see storm, waveswell 264, 265, 271, 358, 362, 366, 377

weatheringchemical 3, 4, 45, 82hydration 150, 187insolation 144, 149

moisture 149–50salt weathering 150, 187

thermal expansion 150weddellite 204Wexford Harbour, Ireland 285–7, 285, 286windfarms see offshore windfarms

yardangs see arid environments, yardangsYellow River, China 91Yellow River Delta, China 235Yellow Sea 362

zinc 5, 15, 27, 81, 81, 92, 98, 99, 100, 118, 124,130, 131, 184, 194–5, 194, 196, 204, 204,205, 206, 209, 210, 216, 217, 219–20, 219,237, 253–4, 254, 340, 371, 383

zooxanthellae algae 304Zostera 247

ESD02 9/5/06 3:10 PM Page 441

environmental sedimentology

Environment

kevin kevin

chris chris

mountain environments

arid environments

fluvial environments

urban environments

lake environments

estuarine environments